Proceedings of the
Edited by W. N. Everitt
Editor
W. N. Everitt Department of Mathematics University of Dundee Dundee D1 4HN Scotland
AMS Subject Classifications (1980): 33 A10, 33 A35, 33 A40, 33 A45, 34Axx, 34 Bxx, 34C15, 34C25, 34D05, 34 D15, 34E05, 34 Kxx, 35B25, 35J05, 35 K15, 35K20, 41A60 ISBN 3540102523 SpringerVerlag Berlin Heidelberg New York ISBN 0387102523 SpringerVerlag NewYork Heidelberg Berlin This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, reuse of illustrations, broadcasting, reproduction by photocopying machine or similar means, and storage in data banks. Under 54 of the German Copyright Law where copies are made for other than private use, a fee is payable to the publisher, the amount of the fee to be determined by agreement with the publisher. by SpringerVerlag Berlin Heidelberg 1980 Printed in Germany Printing and binding: Beltz Offsetdruck, Hemsbach/Bergstr. 2141/3140543210
This volume is dedicated to the life, work and memory of ARTHUR ERD~LYI
19081977
PREFACE
These Proceedings form a record of the plenary lectures delivered at the fifth Conference on Ordinary and Partial Differential Equations which was held at the University of Dundee, Scotland, UK during the period of three days Wednesday to Friday 29 to 31 March 1978. The Conference was originally conceived as a tribute to Professor Arthur Erd~lyi, FRSE, FRS, to mark his then impending retirement from the University of Edinburgh. A number of his colleagues,
including David Colton, W N Everitt, R J Knops, A G Mackie, and G F Roach, met in Edinburgh early in 1977 in order to make provisional arrangements for the Conference programme. At this meeting it was
agreed that Arthur Erd~lyi should be named as Honorary President of the Conference. A formal invitation to attend the Conference was issued
to him in the autumn of 1977, and this invitation Arthur Erd~lyi gladly accepted, expressing his appreciation for the thought and consideration of his colleagues. Alas, time, in the event, did not allow of these
arrangements to come about; Arthur Erd~lyi died suddenly and unexpectedly at his home in Edinburgh on 12 December 1977, at the age of 69. Nevertheless it was decided to proceed with the Conference; invitations had been issued to a number of former students, collaborators and friends of Arthur Erd~lyi to deliver plenary lectures. The Conference
was held as a tribute to his memory and to the outstanding and distinguished contribution he had made to mathematical analysis and differential equations.
VI
These Proceedings form a permanent record of the plenary lectures, together with a list of all other lectures delivered to the Conference. This is not the time and place to discuss in any detail the mathematical work of Arthur Erd~lyi. Obituary notices have now been
published by the London Mathematical Society and the Royal Society of London. Those who conceived and organized this Conference are content
to dedicate this volume to his memory. The Conference was organized by the Dundee Committee; E R Dawson, W N Everitt and B D Sleeman. It was no longer possible to follow through the original proposal for naming an Honorary President. Instead, following the tradition
set by earlier Dundee Conferences, those n~med as Honorary Presidents of the 1978 Conference were: Professor F V Atkinson (Canada) Professor HW Knobloch (West Germany). All participants are thanked for their contribution to the work of the Conference; many travelled long distances to be in Dundee at the time of the meeting. The Committee thanks: the University of Dundee for generously supporting the Conference; the Warden and Staff of West Park Hall for their help in providing accommodation for participants; colleagues and research students in the Department of Mathematics for help during the week of the Conference; the Bursar of Residences and the Finance Office of the University of Dundee. As for the 1976 Conference the Committee records special appreciation of a grant from the European Research Office of the United
States Army; this grant made available travel support for participants from Europe and North America, and also helped to provide secretarial services for the Conference. Professor Sleeman and I wish to record special thanks to our colleague, Commander E R Dawson RN, who carried the main burden for the organization of the Conference. Likewise, as in previous years, we
thank Mrs Norah Thompson, Secretary in the Department of Mathematics, for her invaluable contribution to the Conference.
W N Everitt
C O N T E N T S F. V. Atkinson Exponential behaviour of eigenfunctions and gaps in the essential spectrum .... B. L. J. Braaksma Laplace integrals in singular differential and difference equations ........... David Colton Continuation and reflection of solutions to parabolic partial difference equations ..................................................................... W. N. Everitt Legendre polynomials and singular differential operators ...................... Gaetano Fichera Singularities of 3dimensional potential functions at the vertices and at the edges of the boundary ......................................................... Patrick Habets Singular perturbations of elliptic boundary value problems .................... F. A. Howes and R. E. O'Malley Jr. Singular perturbations of semilinear second order systems ..................... H. W. Knobloch Higher order necessary conditions in optimal control theory ................... J. Mawhin and M. Willem Range of nonlinear perturbations of linear operators with an infinite dimensional kernel ............................................................ Erhard Meister Some classes of integral and integrodifferential equations of convolutional type ............................................................ B. D. Sleeman Multiparameter periodic differential equations Jet Wimp Uniform scale functions and the asymptotic expansion of integrals ............. 251 ................................ 229 151 131 I]5 83 25 1
54
]07
165
82
Lectures ~iven at the Conference which are not represented by contributions to these Proceedings. N. I. AIAmood Rate of decay in the critical cases of differential equations R. J. Amos On a Dirichlet and limitcircle criterion for secondorder ordinary differential expressions G. Andrews An existence theorem for a nonlinear equation in onedimensional viscoelasticity K. J. Brown Multiple solutions for a class of semilinear elliptic boundary value problems P. J. Browne Nonlinear multiparameter problems J. Carr Deterministic epidemic waves A. Davey An initial value method for eigenvalue problems using compound matrices P. C. Dunne Existence and multiplicity of solutions of a nonlinear system of elliptic equations M. S. P. Eastham and S. B. Hadid Estimates of LiouvilleGreen type for higherorder equations with applications to deficiency index theory H. GrabmUller Asymptotic behaviour of solutions of abstract integrodifferential equations S. G. Halvorsen On absolute constants concerning 'flat' oscillators G. C. Hsiao and R. J. Weinacht A singularly perturbed Cauchy problem Hutson Differential  difference equations with both advanced and retarded arguments
XI
H. Kalf The Friedrichs Extension of semibounded SturmLiouville operators R. M. Kauffman The number of Dirichlet solutions to a class of linear ordinary differential equations R. J. Knops Continuous dependence in the Cauchy problem for a nonlinear 'elliptic' I. W. Knowles Stability conditions for secondorder linear differential equations M. KSni$ On C~ estimates for solutions of the radiation problem R. Kress On the limiting behaviour of solutions to boundary integral equations associated with time harmonic wave equations for small frequencies M. K. Kwon$ Intervaltype perturbation of deficiency index M. K. Kwong and A. Zettl Remarks on Landau's inequality R. T. Lewis and D. B. Hinton Discrete spectra criteria for differential operators with a finite singularity Sonssun Lin A bifurcation theorem arising from a selection migration model in population genetics M. Z. M. Malhardeen Stability of a linear nonconservative elastic system J. W. Mooney Picard and Newton methods for mildly nonlinear elliptic boundaryvalue problems R. B. Paris and A. D. Wood Asymptotics of a class of higher order ordinary differential equations H. Pecher and W. yon Wahl Time dependent nonlinear Schrodinger equations system
XII
D. R a c e
On necessary and sufficient conditions for the existence of solutions of ordinary differential equations T. T. Read Limitcircle expressions with oscillatory coefficients R. A. Smith Existence of another periodic solutions of certain nonlinear ordinary differential equations M. A. Sneider On the existence of a steady state in a biological D. C L Stocks and G. Pasan Oscillation criteria for initial value problems in second order linear hyperbolic equations in two independent variables C. J. van Duyn Regularity properties of solutions of an equation arising in the theory of turbulence W. H. yon Wahl Existence theorems for elliptic systems J. Walter Methodical remarks on Riccati's differential equation system
Address list of authors and speakers N. AIAmood: Department of Mathematics, HeriotWatt University, Riccarton, Currie, EDINBURGH EH14 4AS, Scotland R. J. Amos: Department of Pure Mathematics, University of St Andrews, The North Haugh, ST ANDREWS, Fife, Scotland G. Andrews: Department of Mathematics, HeriotWatt University, Ricearton, Currie, EDINBURGH EHI4 4AS, Scotland F. V. Atkinson: Department of Mathematics, University of Toronto, TORONTO 5, Canada B. L. J. Braaksma: Mathematisch Instituut, University of Groningen, PO Box 800, GRONLNGEN, The Netherlands K. J. Brown: Department of Mathematics HeriotWatt University,
Riccarton, Currie, EDINBURGH EHI4 4AS, Scotland P. J. Browne: Department of Mathematics University of Calgary,
CALGARY, Alberta T2N 1N4, Canada J. Carr: Department of Mathematics HeriotWatt University,
Riccarton, Currie, EDINBURGH EH14 4AS, Scotland D. L. Colton: Department of Mathematics NEWARK, Delaware 19711, USA A. Davey: Department of Mathematics University of NewcastleUniversity of Delaware,
uponTyne, NEWCASTLEUPONTYNE NEI 7RU, England P. C. Dunne: Department of Mathematics HeriotWatt University,
Riccarton, Currie, EDINBURGH EHI4 4AS, Scotland M. S. P. Eastham: Department of Mathematics ~ n r e s a Road, LONDON W. N. Everitt: Department of Mathematics DDI 4HN, Scotland G. Fichera: H. Grabm~ller: Via Pietro Mascagni 7,00199 ROMA, Italy Fachbereich Mathematik, Technisehe Hochschule Darmstadt, D 6100 DARMSTADT, Sehlossgartenstrasse 7, West Germany P. Habets: Institut Mathematique, Universit~ Catholique de Louvain, Chemin du Cyclotron 2, 1348 LOUVAIN LANEUVE, Belgium S. B. Hadid: Department of Mathematics, Chelsea College, Manresa Road, LONDON The University, DUNDEE Chelsea College,
XIV
S. G. Halvorsen:
D. B. Hinton:
F. A. Howes:
G. C. Hsiao:
V. Hutson:
H. Kalf:
Fachbereich Mathematik, Technische Hochschule Darmstadt, D 6;00 DARMSTADT, Schlossgartenstrasse 7, West Germany
R. M. Kauffman:
H. W. Knobloch:
R. J. Knops:
Department of Mathematics, HeriotWatt University, Riccarton, Currie, EDINBURGH EHI4 4AS, Scotland
I. W. Knowles:
M. K~nig:
R. Kress:
M. K. Kwong:
R. T. Lewis:
S. S. Lin:
Department of Mathematics, HeriotWatt University, Riccarton, Currie, EDINBURGH EHI4 4AS, Scotland
M. Z. M. Malhardeen:
Department of Mathematics, HeriotWatt University, Riecarton, Currie, EDINBURGH EHI4 4AS, Scotland
J. L. Mawhin:
Institut Mathematique, Universit~ Catholique de Louvain, Chemin du Cyclotron 2, 1348 LOUVAIN LANEUVE, Belgium
XV
E. Meister:
Fachbereich Mathematik, Technische Hochschule Darmstadt, 6100 DARMSTADT, Kantplatz I, West Germany
J. W. Mooney:
R. E. O'Malley Jr:
Program in Applied Mathematics, Mathematics Building, University of Arizona, TUCSON, Arizona 85721, USA
G. Pagan:
Department of Mathematics, Royal Military College of Science, Shrivenham, SWINDON SN6 8LA, England
R. B. Paris:
H. Pecher:
D. Rece:
T. T. Read:
B. D. Sleeman:
R. A. Smith:
Department of Mathematics, University of Durham, Science Laboratories, South Road, DURHAM, England
M. A. Sneider: D. C. Stocks:
Italy
Department of Mathematics, Royal Military College of Science, Shrivenham, SWlNDON SN6 8LA, England
C. J. van Duyn:
W. H. von Wahl:
Universitat Bayreuth, Lehrstuhl fur Angewandte Mathematik, Postfach 3008, D 8580 BAYREUTH, West Germany
J. Walter:
Institut f~r Mathematik, Universit~t Aachen, 51 AACHEN, Templergraben 55, West Germany
R. J. Weinacht:
XVI
J. Wimp:
A. D. Wood:
Department of Mathematics, Cranfield Institute of Technology, CRANFIELD Bedford MK43 OAL, England
A. Zettl:
E X P O N E N T I A L B E H A V I O U R OF E I G E N F U N C T I O N S AND GAPS IN THE E S S E N T I A L S P E C T R U M F.V.Atkinson U n i v e r s i t y of Toronto I. Introduction. In this paper we obtain c o n d i t i o n s on the c o e f f i c i e n t s in c e r t a i n s e c o n d  o r d e r d i f f e r e n t i a l e q u a t i o n s w h i c h yield c o n c l u s i o n s r e g a r d i n g the spectra of a s s o c i a t e d d i f f e r e n t i a l operators. Such results are p a r t i c u l a r l y w e l l  k n o w n for the case y" + (I  q)y = 0 , 0 ~ t ~ ; (i.i)
and its v e c t o r  m a t r i x a n a l o g u e y" + again over (IW  Q)y = 0 , y is a columnmatrix, and (1.3) W, Q
(0,o~), w h e r e
W(t)
h e r m i t i a n and p o s i t i v e  d e f i n i t e .
intervals n e c e s s a r i l y contain a point of the e s s e n t i a l spectrum. As a typical result in the first vein we cite that of ~nol ( 3, p.1562), that for r e a l  v a l u e d ( a n'~ bn) q(t), for w h i c h there
with ~ 0 , (1.45)
(i.i) c o n t a i n s the
As to the second kind of result, we cite (i.i) o b t a i n e d in the early paper of c o v e r i n g also
(I0); c o n s i d e r a b l e extensions,
(4,5,6). in w h i c h we argue
We exploit here a l i t t l e  u s e d method, s u c c e s s i v e l y between: (i) h y p o t h e s e s on the coefficients, cases over sequences of intervals, (ii) the e x p o n e n t i a l g r o w t h or decay, the h o m o g e n e o u s equation,
if any, of solutions of
@(A ), which in some sense measures spectrum. taking the general space functions f(t) case such (1.6) T defined by , (1.7) of locally
We make this last aspect precise, (1.3). We consider measurable that " 2 ilfll d~f I f*(t)w(t)f(t) With (1.6) we associate dt complexvalued columnmatrix
in a hilbert
< co operator
a minimal
(Tf) (t) = wl(t) (Q(t) f(t) with domain f(t) p (~) D(T)
 f"(t))
with compact
We then specify
is the largest number with the property lim inf ii(T {fJ I )fn II ~ ( in D(T) fnx 0 ~(k ) ~ )'
for every
sequence
such that,
i fnl I = 1 , In practice
in (I.I) or
is the
singular
in that we do not argue we see the seems due to (ii) come under
(iii), but make a detour via into two parts, and bring
involved,
with stability
(see the text of Glazman ( 8 , 181183). Implications of the type leading from (i) to of stability The basic y = of E(t) idea is to assooiate,~ (1.3) a function i y*Wy ,
with a solution
y*'y' +
of, roughly
speaking,
Lyapunov
on this topic.
(ii) and
b o u n d e d below,
0(e
~t) (see(8)
for the case of the m u l t i d i m e n s i o n a l Laplacian. case the conclusion is due to P u t n a m a s s u m p t i o n s that > q(t) ( i~
course be squareintegrable,
function a s s o c i a t e d w i t h some initial b o u n d a r y condition. We remark in passing that the c o n d i t i o n b o u n d e d b e l o w ensures that at ~ (i.I) that q(t) be
need similar more general c o n d i t i o n s w i t h the same effect, that the e s s e n t i a l spectrum will be nonvacuous, W e shall w e a k e n the p o i n t w i s e q above to integral analogues, (2);
and p( I ) finite.
Brinck
an e x a m p l e c o n s i d e r e d r e c e n t l y by H a l v o r s e n
( 9 ), who p r o v e s
also an interesting result in the converse sense, w i t h o u t b o u n d e d n e s s restrictions, of solutions of orders then ~ namely that if there exists a p a i r O(tki/2), 0(tk+i/2), where k~ 0.
( I ), and so on the e s s e n t i a l spectrum. We use the symbol * to indicate the h e r m i t i a n c o n j u g a t e as in (1.6), (i.ii). For a c o l u m n  m a t r i x
S
t r a n s p o s e of a matrix,
Ifl = 4(f'f);
for a square m a t r i x
subject to I ; we w r i t e
be d e n o t e d by
Re Q = (Q + Q*)
In
~ 2
differential
inequality,
for w h i c h we need onesided over an interval;,this of a result which, in (I) use this inequality with the hypothesis exponential of intervals. as a foundation that in the
restrictions
on the coefficients
forms the secondorder m a t r i x analogue 2nth order scalar case, was used criteria. In ~ 3 combined we for limitpoint ~(k ) ~
behaviour
We then specialise
to make them ~hold over all intervals of a certain length, as to obtain results of a known type on the exponential behaviour, (~) > including pointwise 0; our assumptions (or q(t) real). In behaviour, on ~5 Q(t)
the effect that onesided restrictions intervals, abstracted ~,o~), of imply a certain degree of exponential we develop more general criteria, spectrum to contain intervals, between the degree of exponential the magnitude as to obtain order results as they overlap, provide hypotheses. Acknowledgements: discussions It is a pleasure a different for
behaviour over sequences ~ 5, for the sequences of arguments so ; in so far (5, 10), but on the
and in ~ 8 C (~)'
and stability
the results
agree w i t h those of
helpful
with Prcfessors
W. N. Everitt,
W. D. Evans,
to take part in the 1978 Dundee Comference on Differential is also made to the continued through support of the National Research Council of Canada,
2.
The basic inequality. The following result is very similar to Theorem 1 of where the scalar 2nth order case was dealt with. Subject W , Q we have hermitian [a, b] let, for positive constants
(1),
A 1 and A 2 , and for a continuously differentiable matrix H(t) , there hold the inequalities (ba)2W(t) ~/ AII, Let the columnmatrix and write = max (Re I , 0) , Re Q(t)~/ z(t)
z" + ( ~ W
 Q) z = 0 ,
C z*Wz dt ,
(2.7) (2.8)
C = 2400! ~ A 1 1 + A1 2 + AI2A22) .
In the proof which follows, all integrals will be over (a,b); the differential dt will be omitted. We note the estimates 0 ~ v ' ~ 3(ba)i/2, Iv"l i 6(ba) 2 (2.9)
We have first that the left of (2.7) does not exceed All(ba)2 ~ (zv" + 2z'v')*(zv" + 2z'v')
2All(ba) 2 I (z*zv"2 + 4(z*'z'v'2)) ~_ 72AI2 I z*Wz + 8All(ba) 2 ~ z*'z'v '2 in the (2.10)
required bound (2.78). It remains to deal similarly with the second term on the right of (2.10). By an integration by parts, and use of (2.4), we have Iz*'z'v'2 = ~ z * ( ~ W  Q) zv'2 2~z*z'v'v" (2.9) we have  ~ z*H'zv '2 +
(2.12),
I zv2 z.
z.zv
211)
Integrating by parts again, we have ' 2 , [z*H'zv = 2 Re~ z*Hz'v '2 + 2 ~z*Hzv'v"
(i/4)[ z*'z'v'2 + 4 I z*H2zv'2 + 18A2(ba)4~z*z (2.12) Turning to the last term in (2.11) we have, by (2.9), [z*zv"2 Combining ~ 36(ba)4 Iz*z <~ 36(ba)2Ali I z*Wz .
(2.1112) we obtain
(lj2) Iz. z v 2
(ba) 2 ~ z*Wz .
(2.13)
(2.10), yields a bound of the form (2.78). holds, of course, if v is replaced by lv. z .
We supplement the above with a pointwise bound for Under the conditions of Lemma 1 we have z(t)v' (t) = ~(z'v' + zv")dT , v~ and so, by a vectorial version of the CauchySchwarz z*(t) z(t)v'2(t) ~ (ba) I (z'v'+zv")*(z'v'+zv")dt
inequality,
Using the above estimations we get z*(t) z(t) ~ (ba) 5(ta)2(bt)2(A3 ~ + A4) ~ z*Wz at, (2.14)
where A 3, A 4 depend only on AI, A 2 . Again, similar results were proved in ( 1 , p. 170) for the 2nth order scalar case. 3. Exponential behaviour. We now suppose that the hypotheses of Lemma 1 are satisfied over a sequence of intervals 0 ~a I ~ bI ~ a2 < b2 ~ (ar, br), with . . . . (3.1)
with the same constants AI, A 2 throughout. We will suppose also that 1 is such that ~( I ) ~ 0. We take some nontrivial solution y of (1.3), and write
Wr = Let ~i u
(3.2)
function
rs (t)
Vr(t)y(t)
Urs(t) = (i  Vs(t))y(%),
By the definition of ~! i ) we have that, for given and sufficiently large r , say for. e ~ r O, II(T I )Ursll 2 ~/ ~12 ~ ~u*Wurs rs dt
~12 ~
y*Wy dt
>1 rl 2 ~
vanishes in (br, a s) , we have that
w.
(3.3) (T  t )Urs(t)
IIcT
where C
~)Ursll 2 .<
CCWr + w s)
wj
(3.6)
We distinguish
wj
.>
co
(3.7)
wj ~
0.
(3.8)
Thus either (3.7) 6r (3.8) is the case. We next note that it follows from (3.6) that
h__ wj
and so, by (3.4), that
>i
Mlk w m
mk
ro
~/
(Mll)Mlki
wm
(3.9)
(MII)MI kI
convex,
increasing
or else ultimately
decreasing,
m I , kI wm , if
can be represented
nlk o + n2(kol)
, with nonnegative
integral
nl~ n 2 . It then
follows from (3.9) that Wm+k > / subject to m ~ Similarly, k+k I , 2k ~ kI . (3.12) in the case (3.8), we can choose integers (M 1  l)Mlki wm (3.11)
m I , kI
m~
m I, k ~
kI , and
can then deduce from (3.9) that, subject to (3.12), we have Wm_ k ~/ (M 1  l)Mlklwm . ~ = ~ (k), which measures wm  we set (3.14) (3.13)
lira sup k 1 lim sup ! in Wm+ k  in WmI.~ I ~9.~ ~ It follows from (3.1213) that in M 1 = in (1 + ~12C 1)
(3.15)
Our argument may be summed up, with a slight loss of precision, in Theorem 1. Let, in the intervals [ar, br~ AlI, Re Q(t) ~/ Hr'(t), l(brar)Hr(t)l~ A2,(3.16) is hermitian and continuously differentiable, be a nontrivial solution of (1.3). Then, with by (3.14), we have (3.17) ~l { (0,~).
in (1 + ~2(~ )cl).
This follows from (3.15), which holds for any We have discussed the case otherwise trivial. 4. ~>
Corollaries regarding global exponential behaviour. We have now completed an argument to the effect that if
suitable conditions are imposed in sequences of intervals, and is not in the essential spectrum, then the solution exhibits exponential behaviour in an integral sense over this sequence of intervals. We shall later use this formulation to derive results about the essential spectrum. At this stage, however it is appropriate to give some more specialised and simpler formulations, including information on pointwise behaviour. We may work in terms of intervals of fixed length if
10 the coefficient certainly us assume of k has a positive Taking lower bound, case and so (1.3), let
W(t) >/ A l I ,
so that the first of (3.16) For the rest of (3.16) Re Q(t)~/ H'(t, where in H(t, T)
we assume T)
for any ~
, IH(t, T ) I %
A 2, T ~ t
differentiable
t , where 2.
Theorem
Under spectrum
essential
solution
e~ty(t) or else
(4.3)
e 2~t I Y2( c ) d~ t as t 9 oO from Theorem e2~t We obtain (4.3) from (1.1)
~oo
(4.4)
It follows
(4.4)
or else
I Y2(~)d~
(4.5)
>
0 .
(4.5)
, with real
of Putnam,
a pointwise W(t) = I
with
bound other
been considered
(15);
related
Similar
be seen by means
of (2.13) ea~t
(or (4.5))
imply that
o.
pointwise
(4.6)
statement
for
y'(t)
indeed be made in the scalar case (1.1), with real real q(t)
details. We have in any case, under the conditions of Theorem 2, that for some 72> 0 either ~9 0,
(4.7)
t+;
277tI Iy(~)
t
y'(c
)t2
d~
9~
. ~0,Oo ). we
(4.8)
note that some criteria for this situation are almost immediate. Theorem 3. Let (4.1) hold and, with let 9 0 (5.1) W(t) continuously
differentiable,
~ 0
0 is in the essential
(5.2)
spectrum. The hypothesis (5.2) ensures (4.2), and so it is sufficSent solution of (1.3) has the exponential
behaviour implied by (4.7) or (4.8). To this end one considers the growth or decay of
E' = ~y*W'y *
E(t)
Y*'QY + Y * Q Y
proves the result. In particular, we have the conclusion in the case (1,1) if
Sequences of large intervals, We now revert to our main line of argument, in which we
consider the behaviour of the coefficients and of solutions over sequences of intervals, rather than over the whole semiaxis. We suppose that conditions are imposed on the coefficients which limit the exponential behaviour of solutions over a sequence of "large" intervals; within these large intervals we wish to be able to select "small" interval, satisfying the requirements of Theorem 1. The principle Theorem 4. with 0 ~ cI <d I G 02 < .... ?(t) / (6.2) 0 , as t~oo , (6.3) (6.1) is embodied in ~r' drD'
and a positive, continuously differentiable function such that, on the c r, dr , W(t) ~ ~2(t)l ~'(t)/~2(t) and ~ ,
I
G
(t)dt
~ Oo
as
(6.4)
Let
R(t), S(t)
be hermitian on the
continuously differentiable, S(t) on the ~cr, dr~ we have Re Q(t) ~/ R ' ( t ) + IR(t)l ~
S(t)
,
(6.5) (6.6)
K17(t)
and
I
~
I t) dt
I
<~ K2 ~ ~(t) dt
e '
(6.?)
13 where Let K I, K 2 ~= are positive constants. ~(~ ) be such that if y(t) is any nontrivial
solution of (1.3), and F(t) = then for any ~ >0 y*'y' + y*Wy , there is an ~ (~ rI such that, for (6.8) r >/r I ,
+ @)lTdt,
t', t" ~ ( C r , d r ) . ( 6 . 9 )
In (i + ~2DI) where
D  lO 4 ~(~, ~12)o2 ~
Here p is as in (1.8), and D
+ 2K22 t.
of
c6.n)
is a modification
not precise, and are inserted only to make plain their independence of the other parameters. similar in nature to ~ The quantity ~ is
or to the general indices of Bohl. If ~ ~is arbitrary, and (6.10) shows that If ~
0, we choose ~
as to obtain the best upper bound for as order of magnitude is concerned. We use Gto determine the
, at least so far
for some infinite sequence of kvalues, f 2 k o   ~ ~ 7 d t ~ 2(k+l)O ~ , c~ We determine 2k intervals (Ckr, Ck,r+l), Ckl = ck , such that ~k~,~
7 d r   ~ , r = 1 . . . . . 2k. (6.13) ckw It then follows from (6.7) that at least k of these intervals
must satisfy
14
~< 2K2G.
(6.14)
a sequence of values satisfying (6.12). This process yields an infinite sequence of intervals (am , bin), which are to be am 9 oo as m 9 oo ,
continuity of y
~ ' am % t
(6.16)
sup in 7 (t)  inf ln~ (t) and so 7(am) exp (o~) ~< 7(t) Hence, by (6.16),
~o~
~ ~(am) exp(6(~ ).
(6.17)
(binam) ~ (am)eXp(6~) ~~<(bmam ) V (am)eXp(~6)' We next calculate admissible values of For AI we need, in view of (6.2), that A I <~_ (bmam)2~2(t) and so, by (6.1718), we may take AI = o2exp( Passing to the remainder of (3.16), we take t Hm(t) = R(t) + ~ S(~)dC, so that, by (6.6), (6.14) and (6.16), ]Hm(t)) ~ R(t) + IO2 (~)I~I(~)S(Y~)IdxZ ~(am) exo(O~) (KI + 2K2O" ). , am <~t ~ b m , A I, A 2
(6.18)
in (3,16)..
15 Hence (bma m) IHm(t)l~<oexp(2~~)(K 1 + 2K2~) := A 2 . We insert these values in (2.8), and obtain a value (6.19)
Noting that
may have any positive value in (6.15) if we see that for large m we
We now consider (3.14). We must first consider the variation of (bm  am ) with m . We claim that 9 1 (6.20)
I(bmk  am+k)/(b m  am)l 1/k as m, k ~ o~ Ck+ 1 ~ Write now sup I?'(t)/?2(t)l, am < , subject to bm <~ ...
(6.21)
Ck+ l <
t~
dk+ 1
Using (6.17) with notational changes we have lln (bm+ k  am+k )  in (bm  am)l~ linT(am+ k )  in ~ ( a m ) l + 2 o ~
by (6.12) and (6.21). This proves (6,20), in view of (6.3) Suppose now that We have trivially that wm ~ (bm  am ) max F(t) , (6.22) y is a nontrivial solution of (1.3).
where the maximum is over ~am, b ~ . opposite sense we denote by (am, bin). We have
(o4 m, ~ m )
~ I F(t)dt g
wm +
(6.23)
numerical factor, by the expression in the braces in (2.13). Hence, by (6.23), 31(bm  am ) min F(t) ~ Wm(1 + C l) . (6.24)
It follows that, subject to (6.21), lln Wm+ k  in Wml ~ sup in F(t)  inf in F(t) + in (3+3C l) +
+ lln (bm+kam+k)  in (bmam) ], where the "sup" and "inf" are over (Ck+ l, dk+l). In view of (6.20), it now follows from (3.14), (6.9) and (6.12) that
We now get (6.10) as a consequence of (3.17). This proves the result of Theorem 4. 7. More on the case that ~0,o~) is in the spectrum.
We go back to the topic of ~ 5, to generalise the approach there given. Whereas in 95 we needed that the solutions should
not grow or decay exponentially when considered over the halfaxis, we can now exploit the situation that this is the case over a sequence of "large" intervals. As indicated after the statement of Theorem 4, it is a question of imposing extra hypotheses so as to ensure that ~= 0 in (6.9). The nature of
these extra hypotheses will depend on the stability argument being used. With the technique indicated in (5.3) we have W(t) continuously
~21W'1dt = o ~ ~ ~ dtl !
(7.1)
(7.2)
17
Suppose that ~ is real and positive. equally use
if
In (6.9) we may
E(t)
instead of
~'F.~~l~t
This follows from (1.11) satisfies E(t) ~
= of~,dt
I.
(7.3)
as given by
ly'(t)l 2 + ~ ~ 21y(t)} 2
(7.4)
!!q(t)Idt
~ 0 , q(t) Q
(7.5) be realvaluedl
and dispense with the requirement that it was not required in Theorem 5 that
be hermitian.
In a second example, we work in terms of an integral of Q(t) take rather than its pointwise values. For simplicity, we now Q(t) hermitian. R(t)
Theorem 6. Let (6.14) held, and also (7.1). Defining in the ~c r, dr~ by
t
R(t) = where Q ~ Q(x= )d~ ~ , Cr~ t ~ dr ' (7.6) is hermitian, let R(t), as t ~ oo W~(t)R(t)W~(t) = o ?(t) 0,0). (7.67)
The proof is similar to that of the last theorem, using this time the Lyapunov or energy function El(t) = yl*y I + ~ y*Wy , Yl = y'  Ry (7.8)
(7.9)
18 We then get an analogous result to (7.3). It is easily seen that E1 may replace F in (6.9) when ~\~ 0, in view of (7.6). q
j iT+l, t
q(t)dt T
as
~ ~o
(7,11)
intervals whose length is unbounded. As a further example, we take the case of (1.2) with w continuously twice differentiable, and q continuous and,
for simplicity,
energytype function 2 _i 2 A E 2 ( t ) = y ' w 2 + ~ y W + yy,w,w3/2 , ~ for which I E2, = 2yy,qw~ + y2qw 'w" 3 / 2 + y y , ( w , w  3 / 2 ) ' Taking ~ real and positive, and assuming that
(7.12)
(7.13)
lw'w3/ l <
we deduce that
(7.14) l w.w l +
, with (7.16) as t 9 =6 .
o 0
Theorem 7.
X1qrw
, w"w29 0
Then the spectrum includes ~ , c O ) . The proof follows the same lines, replacing by F in (6.9)
19 8. Order of magnitude of gaps in the essential spectrum. In the hermitian case, to which we now confine attention, t (~) gives the distance of ~ is real, ~ from the essential spectrum p(~ ) as
will be essentially the same as that of the function the length of a gap with centre /~ in the essential
spectrum~ it is the latter function which has been investigated by Eastham. In a series of papers (see (6)) he has extended earlier results on the relation between this order of magnitude and the continuity and similar properties of the coefficients, The present method, as we have developed it for the secondorder case, seems to have similar power to the singular sequence method used in ( 6 ), at least in certain cases. It then
becomes a challenge to develop the stability techniques used here so as to cover the remaining cases obtained by the other method. We consider the matrix case (1.3), with hermitian and consider the order of natural choice of assume that ~" t(X) for large real ~ Q(t), .A
in (6.1011) is
I / ~ ( ~ )! here we
~(~ ) ~
immaterial, and
~ ( k ) = O.
"~,
(8,23)
e{X}
_0'{~,~ 1 j
{ x }?.
{8..) {8.5}
~ { X } = ocX> .
We proceed to examine some such cases.
20 As in the case of ~ 7, that in which the positive halfaxis is in the spectrum, one obtains a variety of results according Using E(t), as in (1.11), Q(t),
Theorem 8. Let (6.14) hold. Let also (7.12) hold in the modifed form that "small o" 'big 0 ". Then, as ~ ~oo , on the right is replaced by
(~) =
Using ( 5 . 3 )
0 ( ~ ~) .
in the ~r' dr]'
(8.6)
(8.7)
#(~)
In particular, if and w(t)
O(1)
(8.8)
is bounded above and has a positive lower bound, and q(t) is real
In this scalar case, one may transform to the situation (1.2) by means of the change of variable then given by s(x)p(x) dt = dx/p(x); w(t) is
We next give results which fill in the range between (8.6) and the weakest significant assertion, namely that
p ( ~ ) = o(1).
t
(8.1o)
namely
We now n e e d y e t a n o t h e r L y a p u n o v  t y p e f u n c t i o n ,
21
(8.11)
t
We now treat the case in which and in which W(t) ~(t) is also bounded above. We write lim sup t ~ max IW(t+s)W(t)l. 0 < s < ~ ~ (8.13)
Z (k ) =
Simplifying the situation somewhat, we have Theorem 9. Q(t) Let W(t) be continuous and bounded, and let W(t) also have a
positivedefinite
~ ~ 00 ,
0(~):
E 3' and so
O(x~(~)
O.
~ Y*Wl'Y ,'
(8.1.)
We deduce that
'
_ 0
W(t) is uniformly continuous, W(t) In
from which the result follows. We get the case (8.10) if and a range of intermediate
satisfy a HGlder condition with a suitable exponent. particular, condition, we get (8.6) if W(t)
according to the present method we do not need to impose such conditions for all t , but only on a sequence of intervals
of unbounded lengths. The case (8.10) is among those considered by Eastham ( 5 ), for the case (8.9). For results similar to (8.14) in the scalar
22 case reference may be made to Hartman and Putnam (i0), where the oscillation method is used. As is evident from Eastham's work, a reduction of the order of ~(~ ) depends on more drastic assumptions regarding the
smoothness of the coefficients and, if the present method is used, upon the use of a more developed Lyapunov function. We illustrate this here by obtaining (8.8), the case of bounded gaps, for the scalar case (1.2), with nonconstant w(tl). We use now the function assume w(t) E2 given in (7.12), and
of intervals Kc r, dr~ satisfying (7.16). We have Theorem 10. In addition to (7.16) let w'w 3/2 be bounded on
dtl.
(8.15)
Then
as XA ) 00
sequence of intervals of unbounded lengths. The result of the theorem follows from (7.15) together with the previous arguments. More elaborate Lyapunov functions than (7.12), involving higher derivatives, have often been used to get, stability theorems (13) . However it is not clear how to extend them to the matrix case (1,3), or how to form them in general.
23 REFERENCES I. F. V. Atkinson, Limitn criteria of integral type, Proc.Royal Soc. Edinb.(A), 2. 73(1975), 167198.
I. Brinck, Selfadjointness and spectra of SturmLiouville operators, Math. Scand. 7(1959), 219239.
3.
N. Dunford and J. T. Schwartz, Linear Operators. Part II, (Interscience, New York, 1963).
4.
M. S. P. Eastham0 Gaps in the essential spectrum associated with singular differential operators, Quart. Jour. Math. (Oxford)(2), 18(1967), 155168.
5.
M. S. P. Eastham, Asymptotic estimates for the lengths of the gaps in the essential spectrum of selfadjoint differential operators, Proc. Yoral Soc. Edinb.(A), 74(1976), 239252.
6.
M. S. P. Eastham, Gaps in the essential spectrum of evenorder selfadjoint differential operators, Proc. London Math. Soc.(3), 34(1977), 213230.
7.
W. N. Everitt,
Applicable Analysis, 2(1972), 143160. 8. I. M. Glazman, Direct methods of qualitative spectral analysis of singular differential operators, (Israel 1965).
S. G. Halvorsen, Counterexamples in the spectral theory of singular SturmLiouville operators, Mathematies no. 9/74, Matematisk Institutt, Universitet i Trondheim, Trondheim, Norway.
10. P. Hartman and C. R. Putnam, The gaps in the essential spectra of ~ave equations, Amer. Jour. Math. 72(1950), 849862.
oscillatory case, Proc. Amer. Math, Soc. 51(1975),4954 12. R. M. Kauffman, Gaps in the essential spectrum for second order systems, Proc. Amer. Math. Soc. 51(1975), 5561.
13. A. C. Lazer, A stability condition for the differential equation 193196. 14. C. R Putnam, On isolated eigenfunctions bounded potentials, associated with 135147. y" + p(x)y = O, Michigan Math Jour. 12(1965),
15. D. A. R. Rigler, On a strong limitpoint condition and an integral inequality associated with a symmetric matrix differential 76(1976), expression, Proc. Royal Soc. Edinb. (A),
16. L. B. Zelenko, Spectrum of SchrGdinger's complex pseudoperiodic Diff. Urav. 12(1976), potential,
0. I N T R O D U C T I O N
In this p a p e r we c o n s i d e r singular d i f f e r e n t i a l e q u a t i o n s
(0.i)
xlP dy = f(x,y), dx
and d i f f e r e n c e e q u a t i o n s
(0.2)
y ( x p + i) = f(x,y(xP)).
6 ~n, f(x,y)
is an a n a l y t i c function
po ) w h e r e S = {x 6 ~
> O. Assume
f (x,y)
7 ~6I
b (x) y~), w h e r e I = IN
(x) N
[ k=0
k b k x as x + ~ in S. dy (0.2) b y x p = ~ to e q u a t i o n s for ~
We m a y t r a n s f o r m then
(0.I) and
singular p o i n t of the d i f f e r e n t i a l e q u a t i o n
(0.i) of rank at m o s t p. If D f(x,0) Y I + 0(x 2) as x + ~ then ~ is also a singular p o i n b of the d i f f e r e n c e e q u a t i o n (0.2). The c o n s t r u c t i o n of solutions of (0. I) and (0.2) near the s i n g u l a r p o i n t
o f t e n c o n s i s t s of two parts. I. The c o n s t r u c t i o n of a formal series w h i c h f o r m a l l y satisfies if the formal series for y and the a s y m p t o t i c series for b in (0.3) and For example, (0.1) or (0.2). (0. I) or (0.2)
are s u b s t i t u t e d
26
(0.5)
X c m o
m x
However,
in g e n e r a l
t h i s formal
as x ~ in a c e r t a i n
We shall c o n s i d e r m a i n l y
(0.I) a n d
(0.2), w h e r e
f(x,y)
= A(x)y +
W e a s s u m e t h a t the n x n  m a t r i x integrals. We c o n s i d e r
A(x)
and the n  v e c t o r of L a p l a c e
b(x)
are r e p r e s e n t a b l e
as L a p l a c e
two c l a s s e s
w i l l be d e f i n e d a formal w h i c h has
in sect.
i. We w i l l (0.2),
s o l u t i o n of (0.5)
(0.1) or
as a s y m p t o t i c
expansion
and w h i c h
and b
class A. w i t h the same h a l f p l a n e s 3 (cf. sect. 2 a n d 3). T h e class A 2 of L a p l a c e factorial solution series of expansions. (0.2)
of c o n v e r g e n c e
integrals
integrals
consists
of f u n c t i o n s there exists
which
a factorial (0.5)
(0.i) or
corresponding solution
to a formal
solution
is i m p o r t a n t directly from
series
exists
it m a y be c a l c u l a t e d
In sect.
4 and 5 we c o n s i d e r
is r e p r e s e n t a b l e similar
integral
o f a f u n c t i o n ~(t,y) Also
conditions
for the c l a s s e s A 1 and A2. formal integral solution which has (0.5)
then there
(0.5) as a s y m p t o t i c
Instead of
formal
solutions
ck x o of (0.i) or (0.2).
X k
Xk ~ as k
For these
formal
solutions
a result
similar
to t h a t for
(0.5)
holds. Solutions of (0.i) and (0.2) in the form of L a p l a c e Trjitzinski, integrals have b e e n S i b u y a and (0.i)
Birkhoff,
Horn,
Turrittin,
Harris,
Horn
the d i f f e r e n t i a l
equation
or d i f f e r e n c e
equation
y(x)
= Yo +
S o
expt w ( t ) d t
into a s i n g u l a r
Volterra
integral
equation
for w
(0.2)). Banach
o f this i n t e g r a l
equation
functions
with exponential
bounds
in a sector.
T h i s leads to a
27
solution of Volterra
(0.i) or
integral equations
In sect.
of the results
above.
to a reduction
Our results are related to the w o r k of Horn Malmquist aleo W a s o w [16], T u r r i t t i n [18],
is a polynomial
I. LAPLACE INTEGRALS A N D F A C T O R I A L SERIES We shall consider cases w h e r e the differential and difference b equations (0.I) and (0.2) in
integrals.
integer,
@I ~ @2' ~ ~ 0. Let S 1 = {t 6 ~
@i ~ arg t ~ 82 }
functions ~ such that I i) t i  p ~ 6 C (SI, ~n) and, if @i < @2' then ~ is analytic o
S 1 of S 1 .
in the interior
ii) ~ iii) ~
(~lltl)
m
Let (i.i) Then AI(01, (1.2) G1~ Gi(~) = {x 6 ~ : B @ [@1,@2] such that Re (xPe i@) > g}.
functions f
+ L ~(x), p
where fo ~n and ~ 6 a I (@i' 02' U, P)We now define subsets a 2 (m, ~) and A 2 (~, U) of a I (@, 8, g, i) and ~. Let us agree that a on the set and ana
28
D E F I N I T I O N 2. Let ~ 6 ~, m ~ 0, O =  arg m, H > 0. Let S 2 = $2(~) be the compon e n t of {t E : Ii  e~t I < I} t h a t c o n t a i n s the ray arg t = @. T h e n a2(~,~) is
the set o f f u n c t i o n s 02: $2(~0) Cn such that: i) 02 is a n a l y t i c on S 2(~). ii) 02(t) = 0 (i) exp (HiItl)as t + ~ on s2(w) for all ~I > ~" is the set of analytic
Let G 2 = G2(~I) = {x E : Re(xe i8) > ~}. T h e n A2(m,~) functions f: G2(U) + Cn w i t h the r e p r e s e n t a t i o n 02 6 a 2 (~,~) and fo 6 ~n.
For short we will o f t e n d e n o t e the classes of Laplace integrals A d e f i n e d a b o v e b y AI, A 2 or A 1 (H), A2(H) if we o n l y w a n t to stress the v a l u e of the
p a r a m e t e r ~. Moreover, we w i l l use a similar d e f i n i t i o n for m a t r i x functions. It is w e l l k n o w n (1.3) (cf. D o e t s c h [3, p.45, 174] that f 6 AI(@I, @2' ~' p) implies
f(x) ~ fo + I m= i
F (m) q)mXm as x ~
o n G, then f E A 1 (@I' 82' D' P) for some ~ ~ 0. If f E A 2 (~, ~) then f is r e p r e s e n t a b l e by a factorial series (1.4) f(x) = f o + E m=0 m~fm+ I ~ (~ + I) ... , x E G2(~) , (~ + m)
where f
as t ~ o n (~)
w i t h p = i holds as x + ~ on
factorial
series
(1.4) in an a s y m p t o t i c p o w e r series, c o m p a r i s o n w i t h
formula for the fm+1" A l t e r n a t i v e l y we may w r i t e x m as a factorial series; s u b s t i t u t i o n in (1.3) and c o m p a r i s o n w i t h (1.4) g i v e s also a r e c u r s i o n f o r m u l a for
fm .For the explicit form of this formula c f . W a s o w [23,p.330] In this w a y w e sum a s y m p t o t i c series for functions in A2(~,~) b y factorial series. This is a useful p r o p e r t y since factorial series converge u n i f o r m l y in half planes. This p r o p e r t y
29
tions w h i c h u n d e r c e r t a i n c o n d i t i o n s are a s y m p t o t i c e x p a n s i o n s of solutions in A2(~,~) and c o n s e q u e n t l y m a y be summed to any d e g r e e of a p p r o x i m a t i o n If m > I, then S 2 ( m w ) C S 2 (~) and so A2(~,~) c A2 by facto
rial series.
(m~,~). Consequent(1.4) on G 2
ly factorial series
q91 ~ %92 6 aj
if
~9 I, q02 6 aj
2. THE L I N E A R D I F F E R E N T I A L E Q U A T I O N
We n o w c o n s i d e r the d i f f e r e n t i a l e q u a t i o n
linear and that it is a c o u p l e d s y s t e m of a s y s t e m w i t h a s i n g u l a r i t y of the first k i n d and a s y s t e m w i t h a s i n g u l a r i t y of the second kind. To formulate this we p a r t i t i o n n x n  m a t r i c e s along the n I  th row and column w h e r e 0 < n I < n:
where
Mj h
is
an
n.
x n h matrix,
n2 = n
n 1.
A corresponding
partitioning
of
vectors f =
(2.1)
xlP
d__yy A ( x ) y + b(x) = dx
w h e r e p is a p o s i t i v e integer, and c o n c e r n i n g A and b we assume either case I : A, b 6 A I (81' @2' ~' p) or case 2: p = i and A, b 6 A 2 ( ~ , ~). T h e n we have r e p r e s e n t a t i o n s
(2.2)
A(x) = A
+ Lp~(X)
'
b(x) = b
+ L 8(x) p
and a s y m p t o t i c e x p a n s i o n s
(2.3)
A(x)
Z
m=O
b m
x m as x
in closed subsectors o f G 1 in case I and G 2 in case 2. We assume 21 Allm = 0, A 12m = 0, b Im = 0 if m = 0,1,..., pl; A 0 = 0, (2.4) A 22 + ptI is n o n s i n g u l a r in S in case j o n2 j "
30
Then
we have
THEOREM
i. Suppose
Xo
x m ~s a formal
(2.5)
y(x)
X 0
as x ~ ~ on any closed subsector of G 1 in case i and of G 2 in case 2. The solution y with these properties
REMARK. In c a s e series 2 we m a y which sum the
is unique.
formal (2.1) solution o n Re (xe l@ ~ c x m to a c o n v e r g e n t > ~ (cf. sect. i).
factorial
satisfies
PROOF.
Let u
NI Z 0
m c x , a partial m
sum of the
formal
solution.
Then
xlP
d_uu = A ( x ) u dx
+ b(x)
 c(x)
where
c 6 A. a n d c 1(x) = 0 ( x  p  N ), c 2 (x) = 0(x N) as x + ~ on c l o s e d s u b s e c t o r s 3 of G.. H e n c e w i t h y  u = v w e g e t x Ip d v = A(x) v + c(x) as e q u a t i o n e q u i v a l e n t 3 dx t o (2.1). So it is s u f f i c i e n t to p r o v e the theorem I in c a s e b h = 0, h = 0, large by integer (cf. .. ., N. W e (1.3)
..., NI
for a s u f f i c i e n t l y
f r o m n o w o n or e q u i v a l e n t l y
(2.2)
m__ 1
(2.6) B(t) ~ Z m=N y of 8m t p as t ~ 0 in Sj, 1 ~h = 0 if N < h < N + p i.
(2.1) w h i c h
is 0 ( x N)
as x ~ ~,
and which
belongs
to
w is of c k a s s A. t h e n 3
(2.7)
xlP
dxdYY L p =
(ptw) , A ( x ) y
= Lp(A0w
+ ~ * w).
Hence
B an~[
With
(2.9)
~ = 
(A o
+ p t
I)is
31
the equation
for w
is equivalent to
(2.10)
w = Tw + ~. on A 0 imply that
The assumptions
(A + p t I) i = diag {pltiInl,
(A~ 2 + p t In2)l}
p t I) I
are uniformly bounded on S.. So if n I > 0 then T is singular in t = 0. 3 We solve (2.10) in a Banach space V N of functions v : S. ~n such that N 3 t P v is analytic
II
(2.13)
_l/lit] e <
l/ in AI(81,
82, p, p)
or A 2 ( ~ , l/). It is clear that V N is a Banach space with definition will be used for m a t r i x  v a l u e d Since b Aj , it follows from Using (2.9), (2.6), (2.11) and functions.
norm
If.IfN. A similar
in Sj.
Next we show that T maps V N into V N. From the assumption (2.4) we deduce that l h 6 Vp, 2h 6 Vl, h = 1,2. Since
(2.14)
tki
~
we see that t p Moreover, (~ * v) I and t if t 6 S. then 3
1 N+L
P (~ *v) are analytic on S. if v 6 V N. J
_N_ I
)l
(2.15)
]]{ (Ao + p t i)  1 ( ~ *
v)}lll N
Similarly
l/llti
I(~ * v) 2(t) l < II all ie
and therefore
  I
  I
IIv II N Itp
* tp
82
(2.16)
I[{(A
+ p t I)I
(a v ) } 2 1 1 N i sup t6S ]
It p (A 22 o
+ p t
I ) 1 ln2
Hence exists a
with
(2.8)
and
(2.12)
we see of N such
that that
T maps
VN i n t o
VN a n d
that
there
constant
K independent
Choosing N > N
o
Now
going
backwards
we easily
y = 0 ( x N) Hence,
as x + ~ in c l o s e d if N > N
o
and
Pl
verify
o f the o r i g i n a l
equation
(2.1),
without
assuming
(2.6),
such
that
1!
t
P w is a n a l y t i c
on S. a n d 3 NI X m= 1 c m F (m/p)
Plltl
w(t)
 i  I tp + 0(tP ) as t
0 in S. , 3
W (t)
= O(e
) as
~ in
S.. 3
implies such
that
w does
not depend
o n N. H e n c e
we h a v e
a unique this
that
(2.5)
holds.
By v a r i a t i o n
of ~i w e
see t h a t
to t h e c l a s s
A. (~).~3
COROLLARY.
(2.17)
A(x)
x  h ~(xP),
b(x):
x  h b~(xP),"
where ~ ,
bh,
h = 0,1 . . . . .
pl,
are
of class A I ( 8 1 , 8 2 ,
~,
I) in case i and of
class A 2 (~,~) in case 2. Then, if 5~ cm x m is a formal solution of (2. i), t h e r e exists an analytic solution y(x) holds as x + ~ in
= X ~  ~" x h ~ h ( X p) where ~h 6 Aj (p) and 0 (2.5)
(2.18)
2  82 +
e _< p a r g x _< ~
 81  e(e
> 0),
where 81 = 82 =  a r g ~ in case 2.
33
This substitute m a y be s h o w n using a rank reduction scheme of T U R R I T T I N [21]:
~T
(~))T
"'''
YpI
(2.19)
M ( ~ ) u + v(~),
where
M(~)
Z M ~m, 0 m
M o
1 p
I
(2.19) power
A1 o
AIo 1
"'"Ao pI ..... " .A I o "A o o and that 0 is e i g e n v a l u e M 22o is [] case to d i a g { 0 , M 22}O w h e r e and the result follows.
From
(2.4)
we may
deduce
rows
of z e r o s
of Mo w i t h nonsingular.
multiplicity So we m a y
is s i m i l a r I to
apply (2.1)
theorem we have p =
In c a s e p = 0 in m a y be t r a n s f o r m e d b y x:
a regular
singular
p o i n t in ~. T h i s o f the
to t h e c a s e
I by dividing
both sides
equation
(2.20)
dy d~ _ xi A ( x ) y
+ xlb(x)
If ~(x)
we have ~
= 0 and I and
so w e n e e d
n I = n, p = transforms
(2.4)
is s a t i s f i e d
(2.20).
r e l a t e d to f o r m a l
series solutions.
3. T H E
LINEAR
DIFFERENCE
EQUATION
Here
we consider
the equation
(3.1)
By means (3.2)
we transform
(3.1)
into
34
(3.3)
y ( ( x p + i) I/p) = A ( x ) y ( x )
+ b(x).
We distinguish
t w o cases.
In c a s e
i we assume:
A, b 6 AI(81, in c a s e
82, p, p) a n d in j:
c a s e 2 w e assume:
p = i, A, b 6 A 2 ( w , p). We a s s u m e
Ao = d i a g (3.4)
{In I' A ~ 2}' blo = 0, A Ibm = 0, b Im = 0 if h = 1,2; m=l, .... i~i ~ Icos 8 I if 81 ~ @ < ~2;
if k 6 ~ k {0} t h e n 2 k ~ i ~ S j ; ~ A22 t  e o
I is n o n s i n g u l a r n2 in d e f i n i t i o n
on 8.. 3 j o f sect. i. T h e n we h a v e
H e r e S. a n d G. are d e f i n e d ] ] THEOREM
2. Suppose ~ c xm is a formal solution of (3.3) a n d Re t is bounded o m above on sj. Then there exists a function y E A I ( 8 I, 82 , u, P) in case i, (3.3) if (xP+l) I/p E Gj and such that _ (3.1)
PROOF: instead
(2.7) we h a v e y((x p + i) I/p) = L (e t w(t)) (x) , if P equation (2.10) w i t h (2.8) a n d (xP+l) I/p 6 G.. ] (2.9) n o w r e a d s
(3.5)
w(t)
(e t I  A )i o and
(~ * w + 8)(t) .
Instead of (3.6)
(2.11)
(e t I  A )I = d i a g o
b o u n d e d on S.. H e r e w e u s e the f a c t t h a t letl ~ as t ~ o n S.. ] ] w i t h t h e s e a l t e r a t i o n s we m a y s h o w t h a t a l l s t e p s in the p r o o f o f t h e o r e m i slight modifications If R e t is b o u n d e d remain valid, and t h e o r e m 2 follows. []
with
b e l o w o n S. t h e a s s e r t i o n
of theorem
valid.
In t h i s
is n o t b o u n d e d in t h i s case.
If A I e x i s t s w e m a y m o d i f y t h a t p r o o f o
for t h i s case.
35
of 0 in S.. Then the solution may be extended to a 3 in S. (cf. sect. 4.3). We may estimate this solution by ] the right hand side of (3.5) since (etI  Ao)I is bounded in a
neighbourhood
of ~ in S.. Applying Gronwall's lemma we get an exponential bound ] for the solution. In this way we get a solution of (3.3) in Aj(p') for some above, since this result
~' > p. We do not present details of the proof sketched is a special However, case of theorem 6 in sect. a result corresponding 5.
bounded below on S. also may be o b t a i n e d by transformation 3 ~(x) = z(lx). Then ~(x p + i) = AI(x e ~i/p) ~ w h i c h is of the same type as A2(~, (xp)  Al(x e ~i/p)
ix e~i/p)
i
(x e~i/P),
'
b(x e ~I/p)
6 A 1 (e I + ~, 82 + z, p, p) or A2(~,
1.1).
Hence we deduce
from theorem 2 :
Assume
bounded below on s
(2.3) as x ~ on G. and (3.4) holds in case j, j/= 1,2. Assume R e t i S ] and ~ c x m is a formal solution of (3.3). Then there exists ] o m a function v E A (6 g~, ~, p) in case I, v 6 A~(,~, ~) in case 2 such that = v((xPl)l}p)Isat~sfies (3.3) if (xPl)I/PZ6 Gj and (2.5) as x  on Gj.
y(x)
~) in case 2. Then,
if
ts a formal solution of (3.3), there exists an analytic solution m pi y(x) = h~0 "xh ~ h (xp) where ~h 6 Aj and (2.5) holds as x ~ in (2.18) where
81 = 82 =  arg ~ in case 2.
4. THE N O N L I N E A R D I F F E R E N T I A L We consider
EQUATION. equation
the differential
(4.1)
36
G 3
as in d e f i n i t i o n
j of sect.
I. A s s u m e integer, p = I in
HYPOTHESIS
H. (j = 1 o r 2). L e t Po > 0, p > 0, p a p o s i t i v e ] c a s e j = 2. W e h a v e (4.2) f(x, Y) I= f(Y) fo(y) and t + {Lp~(y, t)} (x), if in ~
(x, y) 6 G.] x A
(0; po ) , = 0 and
where
(4.3)
qD(y, t) ~
as t + 0 in S 3 is a n a l y t i c in A (0; po ), m = I, 2, ....
uniformly Moreover,
on A
(0; Do).
Here ~m(y)
if ~I > Z
a constant
K depending
o n Pl s u c h t h a t
(4.4)
I~(Y,
t) l ! K It p we assume
I exp either
(~lltl)
on ~
(0; DO) x S O . O H I or h y p o t h e s i s H 2. If
In t h e f o l l o w i n g = (\~i' "''' ~n ) 6
hypothesis
1 = ~n
we d e n o t e
~%0
(y, t) = ~Ivl%0(Y' t)
and similarly
for ~
f (y). o
j = I or 2 and in case j:
= diag {O n 22 , A } 22 A + p t I is n o n s i n g u l a r o n S. 3
= Dfo(0)
'
i II,~ 1
(4.5) {~ %0 (0, t)} 1 = 0 ( t p ) a s t + 0 in S. if ] IvI < p ,
{~
Suppose
fo(0)} I : 0 if l~l ! P
(4.1) possesses
that
number ~' > ~ and an analytic solution y o f (4.1) suchlthat y E A 1 (81 , 8 2 , p' , p) in case i and y E A 2 (~, ~') in case 2 and such that closed subsectors of G.. This solution is unique. 3
The proof will be given with estimates, derived, in sect. in s e v e r a l steps: in sect. 4.1 w e g i v e s o m e l e m m a s to (4.1) w i l l be equation exists in (2.5) holds as x + ~
on
4.2 an i n t e g r a l
equation
equivalent
in sect.
4.3 w e s h o w t h a t a s o l u t i o n o f t h i s i n t e g r a l 4.4 t h i s s o l u t i o n
o f t = 0, in sect.
will be extended
to a
in S. a n d in sect. 4 . 5 w e e s t i m a t e ] o f (4.1).
the solution
37
4.1.
SOME
ESTIMATES
that
(4.6)
~ m=0
PROOF. The
2~i
m~O
(0 +)
I
es
(~)ml+hI
ds
2~i
(0 +) I e~S ( l  s  1 )  i s  h as .
In the last integral w e choose as p a t h o f i n t e g r a t i o n a loop e n c l o s i n g the n e g a t i v e axis and the p o i n t s s = exp (2gni/1), g 6 ~. The r e s i d u e in s = i gives (4.6)
(O;p),where
(4.7)
where
I Z ( t ) i~ M I t l 11 exp ( ~ l l t l )
M > O, 1 > O, ~i ~
if t 6 S.(p), ]
(4.8)
IZ*~
(t)
I

<
if
t S.(p),u 6 I,
J
v2 f a c t o r s z 2 . . . . . . .
and
~ O, where z * v i s
n
the convolution o f
factors z .
n
pROOF: T h e p r o o f e a s i l y follows b y
i n d u c t i o n using
(2.14).
(4.9)
Then
v6I
~, ,~+ o.
Sj (po) , analytic
w~O
38
(4.10) vEI
X d~Iz*~(t)
(4.11) P
{ X z*V(t) d 9EI w
} =
X ~El
d (L z) v w P
on {x E ~: Re (xp ei8) > Pl + kl} where the path of integration in the Laplace
integral is arg t = 8.
PROOF:
ep21tl
, if t E S, (p), ~ 6 I,
* *~ (t) z
is analytic in S (p), ]
(4.13) ~EI
I~9 Z*~ (t) I  K'O K2 k21h max ([k2tl hl, elk2tl) e71tl * <
~ = max (PI' P)' k. = (Mn r(i)) I/i and K' is a constant dec z z Pl
i. In case p = ~ we have
(4.14)
Lp(~EiX ~
PROOF.We use lemma 2. From (4. 12) and (4.8) we deduce (4.15) ID9 z*W (t) l < *

K2
r(l~ll)
I(tT) hI e
u21tTITI~III
h+]~ll1 B(h,I~ll).
89
4.2. REDUCTION OF THE DIFFERENTIAL EQUATION TO AN INTEGRAL EQUATION From hypothesis H. we deduce ] (4.16) fo(y) = [ ~6I b y , (y, t) = <0 [ By(t) Y ~6I ,
v+o
if y 6 A ( 0 ; P ) , o (4.17) b t 6 S~ and ] 1 SVfo (0) , ~ ( t )
= m~ ~v q) ( o , t ) i f 1 ~ 6 I, t 6 Sj .
From Cauchy's formula for derivatives of analytic functions and hypothesis H. we 3 deduce (4.18) l~i Ib I ~ KoPo[~l , l~(t) l KoP ~ 1 i ~lltl It~ 1 e ,
Let N 6 IN, where N will be chosen later on sufficiently Define UN(X) = NI X i cm x ' zN(t ) = m NI X 1 m  i cm {F(m)} tp I p
Moreover
(4.19)
x Ip
is a formal solution of
gN(x) = 0(xPN),
(4.19) and gN =
(4.17) we may infer with lemmas 3 and 4 Lp YN'  YN (t) = PtZN(t) + ~6ZI { ~ * z*VN + bY zN*m} (t).
(4.21)
These lemmas also imply YN 6 aj(~ 2) for some ~2 > ~ " From
N N
2
(4.20) we deduce
(4.22)
(t)
= 0(tP),YN(t)
O.t{ p
) as
t +0
in
S.. ]
40
(4.23)
xlP
+ v)
 f(x, UN(X))
 gN(x).
w 6 J(v') 3
> ~2 iff
(4.24)
ptw(t)
= AoW(t)
I~I>2
First we remark that H(z, w) exists for some ~"
~+0
in S~ if z, w 6 aj(~') and that H(z, w) Ea=3 ~'') ] > z' on account of lemmas 3 and 4. Moreover, these lemmas imply that (x) = f(x, ~N(X) + v(x))  f(x, U N(X)) on Gj( ~" ). Hence
Next we (4.26)
H(z, w) = ~(z,
I~I>2
where
= DqO(0,.)
{ ~! 3 m ID
vEI
8 (t) +
(v+O) {b +O z *O + O
8 +g ~ z*O} (t) .
(4.7)
~+o
J~I+Iol .
, 1 =  a n d a suitable
189(Z N, t) I ~ M 1 ( ~ ) I W I I t P
of
lemma 4 shows that the series in compact ~ets in Sj, and so (4.26)
convergent on
41
in a Banach
in S.(c) 3
Tw = (A
~N = (Ao + ptI)
Using
(4.22)
and
(4.5) we deduce
~N 6 V N, so in particular
~N 6 VN(). in the
ICll {F(pl)} I +i, where c I is the first coefficient of (4.1). Then i itp i o n S <6 )
3 O
for some sufficiently for w 6 VN() (4.33) In particular of (4.32). We first estimate
small positive
depending i
on N. We consider
H(z, w)
and is sufficiently
a and B
defined by
(4.27)
on S (4.5)
tP
18~(Z, t) I <__ M 3
(Q~) I~I
if z 6 V(),
t 6 Sj(a) N
Now choose
I{(Ao + ptI)I
w(t)}ll
Itp to
z 6 V(g).
42
* w(t)}21
< K'
I I wl IN [tl p
there exists E2,
independent
of z, w and t. Hence
* W 1 IN
<
l lwl IN ,
211~NI I , where the norm is the norm in VN(E 2) N Hi(z , w) = H(Z, w)  ~(z,.) (4.34) . w.
w 6 VN(E), independent 2N
of z, w, t , g. Choose M4 < i 4
(Ao + ptI)ll
= {w 6 VN(E):
 (4.37)
that
1 _
(4.38)
H(ZN + W, W2
exists a unique
solution of w = Tw, so of
(4.24),
in D(E).
OF THE SOLUTION w of
the solution
is known on S.(p)
Choose to 6 S.(p)3 with 21P < jtol Here S + t denotes o sets. We transform decomposition analytic
3 = S~+t3o"
the integral
equation
of u . v for scalar
functions
in its interior:
43
(4.39)
(u ~v)
(to+tl)
= {U(to+
.)~ v}
(tl) +
{V(to+
.)
u}
(t I) + R(u,v)(tl),
and
[tl,
t o ] belong
to
Sj(p).
Here
R(U, v) = R(V, u) and R(u, v) only depends Using induction ~k we may show
(tl) +
kI E j=l
{ *(klj) v
* R(v, v~J)}
(tl),
if k >
that
for an nvector k 6 I, I kl
z whose components
above and
> 2 we have n I
(4.41)
~(z,
tl)m
zmk(t
+ tl) 
~=~
] I
{z.(to+
]
Zl ) (tl) ~J
Here e. is the nvector whose components are zero except for the jth component ] which is equal to one. From the definition of R and ~ it follows that ~ ( z , t I) is determined on S? by the values of z on S.(p). ] ] If z satisfies (4.7) with ~I = o and 1 = 1 on Sj(p)
(4.8)
and
ki+
2 2(Mp)
....
k1
if t I 6 S. Combining these formulas with (4.41) and (4.8) we see that there ] exists a constant Mo independent of t I and z but depending on p and M such that
44
(4.42)
]~(z,
, t I C Sj .
Now we transform of
the operator T into an operator ~ on $7. Let w be the solution ]n (4.24)of w = Tw on S.(P)3 (cf.(4.30)). If z 6 C (S~, (~) then we define
(4.43)
(~z)
I) I {~ ~ z(tl)
+ ~
(tl)}
if t I 6 S~. and
n , ('~e,)
* z(t I) = ~(ZN,.)*
z(t I) +
[ V61
[ j:l
~j {8 (ZN,.) * w v
I~I>_2
* (~e)
+ b w
] }
(t I) , zj
.) ,w)(t I)  YN(to + t I) +
E v6I
{Sv(ZN,
.) , Rv
(w,.) +
(z N, to+ .) * w *v +
I~I>_2
+ R(Sv(z N, Using lemmas 2 and 4, ), W *v) + b v R (w, .) }(tl). (4.28), (4.42) we may deduce that ~ and ~ exist and are
continuous on $7 and analytic in ($7) . These functions only depend on the values 3 ] of w in S. (p). 3 The definitions of T and ~ in (4.30) and (4.43) and (4.39)imply (~) (to+t I) = {~ W(to + ")} (tl) if t16 $73 n (sj(p) to). Hence w (to + .) is a equation
integral
z = ~ z has a unique solution in S7 w h i c h is analytic in (sT) O and continuous 3 3 on S~, this solution is a continuation of w(t + .). Denoting this continuation o J also by w(t + .) we see that w = T w on S.(O) U S < b e c a u s e of the relation of T o 3 3 and ~. Varying t o we get a unique solution of w = T w on S_3 (3),z hence on S.. 3 Thus we have shown that (4.24) has a unique solution w in S. w h z c h is continuous o ] on S. and analytic in S.. 3 ]
4.5. E X P O N E N T I A L
ESTIMATE
FOR THE SOLUTION (4.24) on S,. Let 3 ItI: p}, if p > 0 . for (A + ptI) o i from
the solution w of
45
(2.12)
for H(ZN,
w)
from
(4.26),
(4.28)
and
(4.31)
and
T h e n w e see t h a t t h e r e
exists positive
@ > 1 we have
(4.45) where
g(p) <
(Tlg)
(p) ,
( ' r i g ) (p)
= M {eP3 p
+ ~2
m:2
d m *m (p) g
d m (p~  1 e P B P ) :~ g~m( p } }
By c h o o s i n g
p [0,1]}we
Following
Walter
[23,
p . 1 7 ] we f i r s t
solve v = Ttv.
If u = Llv, 1
then
u(x)
= M(xP3)I
+ t,~ 5m=2
X
m:l
dmum(x)
This equation
has a unique
s o l u t i o n u in a n e i g h b o u r h o o d 1
V of ~ w h i c h
is a n a l y t i c
P) as x p ~ P4 > ~3" T h e n
the halfplane
Re x ~ P4' w h e r e
v(p)
1 (Lllu) (p) = M + 2 ~
;4~ i~
P4i ~
ePX
t h a t v is r e a l  v a l u e d , In p a r t i c u l a r
v(0)
= M
(cf.[3, p.174])
and v(p)
we h a v e g(0)
Suppose
g(po ) = V(Po).
= V(Po)
(4.46)
lw(t) I ~ K .
KO e x p
for some c o n s t a n t
Consequently
LpW e x i s t s
o n Gj a n d
is a solution
of
(4.23).
Therefore
y = UN+
46
(4.47)
y(x)
Ikl=1
w h e r e k = (ko, ..., k Re <j > 0 if j = application THEOREM V). ) 6 ~g+1 (g a p o s i t i v e i n t e g e r ) , K = (i, <i, ..., < ) g g i , ..., g and k . < = k o + klKl + ... + k g < g (cf. sect. 6, c a s e s we h a v e the f o l l o w i n g H I . Let generalization of t h e o r e m 4.
In t h e s e
5. A s s u m e
hypothesis
(4.48)
A m = F(~)
and assume
A ll = 0, A 12 = 0 , b1 = 0 if m = 0 ..... p1 and w 6 I, I~1 + 1,
m (4.49)
~m is n o n s i n g u l a r (4.47) of(4.1) on S I.
A 21 = 0, A 22 + ptI o o
If
(4.1)
has a formal
solution
then there exists a real number" ~' > ~ and on GI(~') (cf. (l.l))such that
an analytic
solution y = L w P oo
y(x) N y
(4.50)
k. K dk x
Ikl=*
as x + ~ on closed subsectors
PROOF. with The proof is s i m i l a r
of GI(V').
to t h a t of t h e o r e m
in o r d e r o f i n c r e a s i n g
of t h e i r r e a l p a r t s to
the s e q u e n c e Then
... and Re Am ~ as m + ~.
t
(4.47) m a y be r e w r i t t e n NI = Y i
x A m . L e t NI zN = Z i 1 m i c {F(m)} I t p m p In g e n e r a l u N ~ A I. W i t h
(4.51)
UN(X)
c m
i x m, U N = L p Z N ,
w h e r e N is c h o s e n (4.19) w e d e d u c e giN(X)
in s u c h a w a y t h a t Re IN_ I < Re A N.
= O(x pIN),
2 gN(x)
I = 0(x N)
and
(4.21).
Instead
of
(4.22) w e n o w h a v e
l
) as t + 0 in S I
TiN(t)= 0 ( t p N) , T N ( t ) = 0 ( t P N 2 w i t h y = u N + v, v = L p W w e d e d u c e (4.23) 
(4.27).
Because
of
(4.51) w e m a y
47
estimate e and 8~ in
(4.28) w i t h I/p r e p l a c e d b y Ii/p. In sect. 4.3 we a d a p t the d e f i n i t i o n of VN(e) b y r e p l a c i n g N / p b y IN/p (cf. (4.29)). In (4.32)  (4.34) and (4.45) we n o w have to r e p l a c e i/p b y
ii/p and in
4.2  4.5 r e m a i n s v a l i d and the p r o o d of t h e o r e m 5 is completed. REMARK. Cases where a formal s o l u t i o n of (4.1) also c o n t a i n s l o g a r i t h m i c terms like
or e q u i v a l e n t l y w i t h (5.2)
(3.2) :
w h e r e f s a t i s f i e s h y p o t h e s i s H. 3 (5.3) (5.4) A o = Df
'
(5,5) (5.6)
119t
~0(O,t)} 1 = 0 ( t (5'.7) ~ fo(O) = 0 if l~I < p P ) as t 0 in S. if 191 < p 3
i m ~' > ~ and an analytic solution y of (5.2) such that y 6 A 1 (e l, e2,v', p) in case I a n d y 6 A2(~o, ~') in case 2 and such that (2.5) holds on closed subsectors of G . 3 This solution is unique.
Suppose
48
reads etwit) In v i e w of
and in
(4.30), and
(4.31) (5.6)
etc.
we r e p l a c e
(A
+ ptl) I by
(A etI) I
91
(3.6), (5.4) the m a t r i x
(A etI) is b o u n d e d on S . ~ A(0;I) o 3 and the m a t r i c e s in (3.7) a r e b o u n d e d on S. N A(0;I) and e v e n on S~ if 3 3 Re t is b o u n d e d a b o v e on S.. Using these m o d i f i c a t i o n s the p r o o f of t h e o r e m 4 3 goes through. D REMARK. Theorem 6 is an e x t e n s i o n to t h e o r e m of a r e s u l t of Harris a~id S i b u y a [7]. 6.
generalization
of t h e o r e m
7. Assume
hypothesis
(5.6) with
of
(5.8)
A 11 = 0, A 12 = 0,if m=l,..., m m
pl;
b um : I
o if
m=o . . . . .
p  1 and ~ ~ ~ , b t + I~
(4.47)
~' > p and an analytic solution y = L w of (5.2) on GI(p') (ef.(l.l))such that P (4.50) holds as x + ~ closed subsectors of GI(P'). This solution is unique.
PROOF. The p r o o f is a m o d i f i c a t i o n 6. of that of t h e o r e m 5, w i t h a n a l o g o u s modifica
tions as in the p r o o f of t h e o r e m
6. A P P L I C A T I O N S In this s e c t i o n we series solutions Finally of first g i v e s u f f i c i e n t and (0.2) exist, conditions in o r d e r t h a t formal m a y be
(0.I)
theorems
applied.
we d e d u c e
a reduction
theorem
differential
equations.
I. A f o r m a l n o n  t r i v i a l
solution
exists
for
(3.3),and
so for
(3.1), w i t h b = 0 if for
(3.4)
A 12 = 0, A 11 is s i n g u l a r P P in the c o n d i t i o n s
(6.1)
y((xP+l)
= X  mc x  m  p , m 1 I/p)  y(x) = X
i
c x m  p
m
~ ~ +
P
X g=l
xPg}. g+l
Formal equations
solutions and
above
(2.1)
(3.3)
the same c o n d i t i o n s
49
to be n o n s i n g u l a r ; treatment of formal
then c
II. R e s u l t s
for formal
s o l u t i o n s ~ c x %m a n a l o g o u s to t h e o r e m s 13 m a y be o b t a i n e d o m of the e q u a t i o n s (2.1) and (3.3) v i a the s u b s t i t u t i o n for ~ is of the same type as that for y and has the
The e q u a t i o n
formal
i contain of
the r e s u l t s
of T u r r i t t i n
[18]
of c o n v e r g e n c e of c o n v e r g e n c e
series which
III. A m a t r i x Assume
(2.1) a n d
(3.3) w i t h b ~ 0 m a y be o b t a i n e d
as follows.
A 12 = 0 and there is no p a i r of e i g e n v a l u e s of A 11 w h i c h d i f f e r b y a P P p o s i t i v e i n t e g e r in the case of (2.1), w h e r e a s in the case of (3.3) we assume the same satisfied and U ( x ) x subsectors for p Ap 11 . If the a s s u m p t i o n s and b ~ of 0 we m a y c o n s t r u c t (3.3) concerning A in t h e o r e m s solution 1,2 41 3 are or (2.1)
an n x n l  m a t r i x
U(x)x ~ of
T h e n we g e t _A 11 P
U((xP+I) I/p)
The r i g h t h a n d matrices
side d e f i n e s
a linear
transformation
T in the space V of n x n l
U. P a r t i t i o n i n g (TU)I(x)
these m a t r i c e s Ul(x)
= {All(x)
ii ii t U1 b N U.  U , N has as e i g e n v a l u e s he Ii p i i p d i f f e r e n c e s of the e i g e n v a l u e s of A . H e n c e p times this t r a n s f o r m a t i o n has P eigenvalue 0 with eigenvector I and no o t h e r i n t e g e r is eigenvalue. A l s o n1 (TU)2(x) ~ A 22 U2 (x) as x + ~ in G . H e n c e all c o n d i t i o n s of a p p l i c a t i o n I are o 3 s a t i s f i e d and the r e s u l t for U follows. Similarly, t h e d i f f e r e n t i a l e q u a t i o n Here the linear t r a n s f o r m a t i o n (2.1) m a y be t r e a t e d (cf.[2]). conditions such t h a t there e x i s t s (4.5) a formal From solution (4.2) and
IV. We n o w g i v e s u f f i c i e n t
are satisfied.
50
f(x, y) : (2 0
A xm)y + [ m ~EI
( [ m=0
b
vm
x m) y~ +
[
m=l
b
om
x m
I~I>S
F r o m this we m a y d e d u c e that there exists a formal solution of o r if n. ~ i) Alllhas P I and no eigenvalue which is a negative integer and (4.1) if n 1 = 0
(6.2)
or ii) A II has e i g e n v a l u e i, but no other n e g a t i v e integer is e i g e n v a l u e of A II , P P (6.2) h o l d s and A 12 p (A22)Ib 2 = b 1 o ol op+l
In these cases we m a y a p p l y t h e o r e m 4. An a n a l o g o u s result (4.5) we n o w assume ii replaced by p A P V. A formal solution H~ and ] holds for the d i f f e r e n c e e q u a t i o n and the c o n d i t i o n s (5.2): instead of Ii P
(5.3)(5.6)
i) or ii) above w i t h A
(4.47) of
(4.1) a s s u m e d in t h e o r e m 5 exists if h y p o t h e s i s
+ kl <I + ...+k
.
<
g
k
k1 +
..
+ kg_
>
..., k
E IN and
then 1 6 {~i
'
"'''
<g}' bl
op+h
0,
b2 oh
Ail ; if P = 0 for

AI2 p
(A22)i o
..., g. We refer to H u k u h a r a
A formal solution
(5.3)(5.6)
is
VI. We m a y a p p l y IV to obtain a b l o c k d i a g o n a l i z a t i o n for linear systems (6.3) x l  P d_yy = C ( x ) y dx [24, t h e o r e m 12.2] and M a l m q u i s t [16, sect. 3 ].
a n a l o g o u s to W a s o w
51
T h e o r e m 8. Suppose C 6 A ( U ) , ]
Let X I
' "
..
'
~
r
be the eigenvalues of
C22 ) o
~ those of C 22
n o "
Assume Z !p (lg  Ih) ~ S 9 if g _< r < h. Then there exists a transformation y : T ( x ) z which takes (6.3) into
(6.4)
C~92(x)}
such that T and ~ 6 A (~' ) for some ~ ' > ~ and ~Ii = cll , ~22 = C22 3 o o o 0 If CI2(x) , C 21 (x) = 0(x N) , then T(x) = I + 0(x N).
P R O O F . F i r s t we substitute y = Q(x)y, w h e r e
ox/to fix t
Then (6.5) (6.3) is t r a n s f o r m e d into x l  P d w = D(x)w, dx 0 iff w h e r e Dl2(x) ~ (6.6)
N o w C II 12 12 22 o Q Q Co d e f i n e s a linear t r a n s f o r m a t i o n in the linear space of 12 r x (nr)matrices Q w i t h e i g e n v a l u e s hg hh, g = I, .... r; h = r + I, ..., n. Hence we may use a p p l i c a t i o n IV w i t h n I = 0 and a solution of A.(~') ] for some ~i > Z" In a similar way we m a y t r a n s f o r m (6.6) exists in (6.4). []
(6.6) to
sector. R e s u l t s of this type m a y be u s e d to reduce linear d i f f e r e n t i a l and d i f f e r e n c e e q u a t i o n s to canonical forms, cf. M a l m q u i s t [16], T u r r i t t i n [18] ,[20].
REFERENCES [i] BRAAKSMA, B.L.J., Laplace integrals, factorial series and singular diffe
series
52
[4]
ERDELYI, A.,
Asymptotic
(1964)
[6]
Integration formelle d'un syst~me des ~quations di~rentiel (4) 19 (1940) 3544.
ordinary differential equations with an irregular type singular point. Ann. Mat. Pura Appl., (4) 82 (1969) 189256.
[15] IWANO, M.,
diff~rentielles dans le voisinage d'un point sin~lier d'ind~termination, II. Acta Math., 74 (1941) 164.
[17] TRJITZINSKY, W.J.,
53
[18]
TURRITTIN,
H.L.,
differential equations in the neighbourhood of an irregular singular point. Acta Math., 93 (1955) 2766.
[19] TURRITTIN, H.L., The formal theory of systems of irregular homogeneous linear difference and differential equations, Bol. Soc.Math. Mexicana (1960) [20] TURRITTIN, 255264.
H.L., A canonical form for a system of linear difference equations, Ann. Mat. Pura Appl., 58 (1962) 335357. H.L.,
[21]
TURRITTIN,
problems, in Asymptotic Solutions of Differential Equations and their Applications. Edited by C.H. wilcox, John Wiley, New York,
1964, 2752. [23] WALTER, W.,
Interscience
David Colton * Dedicated to the memory of my teacher and friend Professor Arthur Erd~lyi
I. Introduction. As is well known, a solution of an ordinary differential equation can be continued as a solution of the given differential equation as long as its graph stays in the domain in which the equation is regular. On the other hand the situation for solutions of partial differential equations is quite different since a solution of a partial differential equation can have a natural boundary interior to the domain of regularity of the equation (c.f.~7]). exceptional circumstances In fact it is only in very
regular solution of a partial differential equation in a given domain can be extended to a solution defined in a larger domain. In the
general case continuibility into a larger domain depends on the solution of the partial differential equation satisfying certain appropriate boundary data on the boundary of its original domain of definition, the classical example of this being the Schwarz reflection principle for harmonic functions. In the past twentyfive years there has been
a considerable amount of research undertaken to determine criteria for continuing solutions of partial differential equations into larger domains and in these investigations two major directions stand out:
* This research was supported in part by NSF Grant MCS 7702056 and AFOSR Grant 762879.
55
i) reflection principles,
by means
both of these approaches have been confined to the case of elliptic equations. The generalization harmonic functions of the Schwarz reflection principle for
pendent variables
a plane boundary was established by Lewy in his seminal address to the American Mathematical Lewy considered Society in 1954 ([20]). In this address
Uxx + uyy
defined in a domain D adjacent on the side y < 0 to a segment o of the x axis. On o, u(x,y) was assumed to satisfy the first order
L/ o), ~(z),
~(z), y(z) and f(z) are analytic in D u o ~ D* (where D* denotes the mirror image of D reflected across o), ~(z) # 0 and ~(z) # 0 throughout D ~ o ~ D * , the variables z = x + iy and the coefficients of (i.i) expressed in terms of
( i . 3)
z* = x  iy are analytic functions of the two independent and z* for z e D V o~ complex variables z
could be continued into the domain D ~ o ~ D* as a solution of (i.I). In particular Lewy showed that the domain of dependence associated
was not valid in higher dimensions, even for the case of Laplace's equation in three variables satisfying a linear first order boundary condition with constant coefficients along a plane. problem of the reflection of solutions to higher dimensional elliptic equations across analytic boundaries was taken up by Garabedian in 1960 ( ~ 2 ~ ) who showed that the breakdown of the This
reflection property is due to the fact that the domain of dependence associated with a solution of an n dimensional elliptic equation at a point on one side of an analytic surface is a whole n dimensional ball on the other side. Only in exceptional circumstances
does some kind of degeneracy occur which causes the domain of dependence to collapse onto a lower dimensional subset, thus allowing a continuation into a larger region than that afforded in general. Such is the situation for example in the case of the
Schwarz reflection principle for harmonic functions across a plane or a sphere (where the domain of dependence degenerates to a point) and the reflection principle for solutions of the Helmholtz equation across a sphere (where the domain of dependence degenerates to a one dimensional line segment  c.f. [41). Such a degeneration can for reflectio ~ analogous
to the classical Huygen's principle for hyperbolic equations, and in recent years there have been a number of intriguing examples of when such a degeneracy can occur (c.f. [9], [21]). The second major approach to the analytic continuation of solutions to elliptic partial differential equations is through
57
This approach
is based on the use of integral operators for partial differential equations relating solutions of elliptic equations to analytic functions of one or several complex variables and has been extensively developed by Bergman ([i]), Vekua ([242) , and Gilbert ([12). The main idea is to develop a local representation of the solution to the elliptic equation in the form of an integral operator with an analytic function with known singularities as its kernel and with the domain of the operator being the space of analytic functions. The problem of the (global) continuation of the solution to the elliptic equation can then be thrown back onto the well investigated problem of the continuation of analytic functions of one or several complex variables. A simple example typical of this approach, and
gations, was that obtained by Erdelyi in 1956 on solutions of the generalized axially symmetric potential equation u xx +u yy +k y
u
(1.4)
if u(x,y) is a regular solution of (1.4) in a region containing the singular line y = 0 and u(x,O) is a real valued analytic function in a yconvex domain D (i.e. if (x,y) e D then so is (x,ty) for l<t<l) then, provided k # 1, 3 . . . . . in D. u(x,y) is a regular solution of (1.4)
/
For generalizations of this result of Erdelyl the reader is The method of integral operators
58 analytic continuation of the solution to Cauchy's problem for elliptic equations which arises in connection with certain inverse problems in fluid mechanics (c.f. [iI], [14]). More recently the author used such
an approach to establish a relationship between the domain of regularity of axially symmetric solutions of the Helmholtz equation defined in exterior domains and the indicator diagram of the far field pattern
([2]).
In contrast to the rather extensive investigations into the problem of continuing solutions of elliptic equations, until recently relatively little work has been done in connection with the corresponding problems for parabolic equations. One reason for this is of course the
fact that solutions of parabolic equations do not enjoy the same regularity properties as solutions of elliptic ~quations, i.e., in general
solutions of parabolic equations with analytic coefficients are not analytic functions of their independent variables. Until about ten
years ago there were (to the author's knowledge) only two rather isolated results on the global continuation of solutions to parabolic equations, both of them concerned with the heat equation in one space variable. The first of these was the fact that solutions of the one dimensional heat equation defined in a rectangle and satisfying homogeneous Dirichlet or Neumann data on a vertical side of the rectange could be reflected as a solution of the heat equation into the mirror image of the rectangle ([25]). On the other hand it was shown by Widder in [26] that
if h(x,t) is a solution of the one dimensional heat equation h xx = h t o' (1.5) Itl < t , then h(x,t) can o
59
Ixl < =, Itl < to, and expressed there as a uniformly convergent series of heat polynomials h(x,t) = Z a h (x,t) nn n=O (1.6)
where the heat polynomials are defined by /~ h (x,t) = ~]: E n k=O x n2ktk (1.7) (n2k) ~k:
This result is noteworthy since it is one of the few cases where it can be stated that eyery sufficiently regular solution of a partial differential equation defined in a given (real) domain has an automatic continuation into a larger (real) domain regardless of what the boundary data is. Note that such behaviour is only true for
analytic solutions of the heat equation, and not in general for classical solutions which are infinitely differentiable, but not analytic, in the time variable. Hence in the development of the reflection and continuation properties of solutions to parabolic equations with analytic coefficients one can expect a different behaviour depending on whether or not the solutions are analytic in the time variable. In this connection it is worthwhile to note that if u(~,t)
is a solution of a linear parabolic equation with analytic coefficients defined in a cylinder~.x(O,T) in ~Rn+l, and if u(~,t) assumes analytic
Dirichlet data on the analytic boundary ~.C~x(O,T), then u(~,t) is an analytic function of its independent variables i n  / ~ x(O,T)(c.f. ~0]). In the following sections we shall survey, with an outline of proofs, some recent results on the reflection and continuation of solutions to linear parabolic equations with analytic coefficients. We
60 corresponding results should also be valid for certain higher order equations. The theory we shall present is far from complete. In
particular we have omitted questions concerning removable singularities as well as the problem of backward continuation of solutions to parabolic equations. Some aspects of this last topic will be
discussed in a survey paper to be presented at the conference on "Inverse and Improperly Posed Problems in Differential Equations" to be held next year in Halle.
II. Parabolic ERuations in One Space Variable. The basic idea behind Lewy's reflection principle for elliptic equations was to develop an integral operator which allowed the general reflection problem to be reduced to one in which the Schwarz reflection principle for analytic functions could be applied. This
is also the main idea behind the method for developing reflection principles for parabolic equations, where in this case the basic theorem employed is the reflection principle for solutions of the heat equation. The operators employed in this analysis are no longer
based on the idea of a complex Riemann function as in Lewy's work but instead on a generalization of the transformation operators for ordinary differential equations developed by Gelfand, Levitan and Marcenko in their investigations of the inverse scattering problem in quantum mechanics (c.f. [7]). To introduce these generalized trans
formation operators we first consider solutions of the parabolic equation Uxx + a(x,t) Ux + b(x,t)u = u t (2.1)
defined in a domain D of the form D = {(x,t): Sl(t)<x<s2(t)10<t<to } and make the assumption that the coefficients of (2.1) are analytic
3* is the reflection of D across ~ defined by D* = {(x,t): s2(t) < x < Sl(t) , 0 < t < to }"
in a domain D to the right of the t axis with lateral boundary x = O. A further change of variables of the form fx u(x,t) = v(x,t) exp { 12 I a(s,t)ds} J0
(2.3)
reduces
(2.1) to an equation of the same form except that a(x,t) Hence without loss of generality we can
where q(x,t) is analytic in D u ~ ~ D* and D = {(x,t):O<x<s2(t), O<t<to} , ~ = {(x,t): x=O}, D* = {x,t):  s2(t)<x<O, O<t<to}. Assume now that u(x,t) in ~ ~ ~ ~ ~*. is a classical solution of (2.4) defined in the form ([3])
where h(x,t) is a solution of the heat equation hxx = h t in ~ problem ~ ~* and E(s,x,t) (2.6)
62 E  E ss + q(x,t)E = E
xx
E(x,x,t)
=  ~
fx
O
t (2.7)
q(s,t)ds
E(x,x,t)
= O .
A solution of (2.7) can be constructed by iteration and is analytic for s2(t)<x<s2(t ) s2(t)<s<s2(t )
'
O<t<t .
O
can be represented
in the form (2.5) follows from the fact that ~ is a Volterra operator and hence I + T is invertible. for parabolic equations various modifications The reflection principle
of the operator
trated in the proof of the following theorem: Theorem 2.1: continuously Let u(x,t) be a classical solution of (2.1) in D, differentiable in D k7 ~, and satisfying + y(t)ut(sl(t),t) = f(t)
~(t)u(sl(t),t)
+ 8(t)Ux(Sl(t),t)
(2.8)
on o, where Sl(t), ~(t), B(t), y(t) and f(t) are real analytic on (O,to). If, for t E (O,to), the non zero vector ](t) = (or always tangent to
(~(t), y(t)) is never tangent to x = Sl(t) x = Sl(t)) and never parallel to the x axis), then u(x,t) ution of (2.1) into D ~ Proof:
o~)D*.
By the above discussion we can reduce the problem to the (2.4)), Sl(t) = O, and assume Furthermore, by solving (to be
63 Hence without loss of generality we can consider the problem of continuing solutions of (2.4) subject to (2.8) with Sl(t ) = 0 and f(t) = O, where u(x,t) is defined in D and is continuously differentiable in D ~ o. We shall only consider the special case when
6(t) # 0 and y(t) # 0 for t c (O,t o) and refer the reader to [5] for full details. In this special case by making a preliminary
change of variables of the form U(X,t) = v(x,t) exp {0 we can assume without loss of generality that ~(t) = 0 and the boundary condition on o is Ux(O,t ) + q(t)ut(O,t) = 0 where q(t) = (2.10) ~(~)dT} (2.9)
Y(t)16(t. )
represented in the form u(x,t) = h(1)(x,t) + h(2)(x,t) + XK(1) ~0 (s,x,t)h (i) (s,t)ds
(2.11)
+ {XK(2)(s,x,t)h(2)(s,t)ds JO where K(1)(s,x,t) is the solution of K (I) _ K (I) + q(x,t)K (I) = K (I) xx ss t K(1)(x,x,t) =  ~
q~)'J
= K (2)
constructed by iteration and are analytic for s2(t)<x<s2(t), s2(t)<s<s2(t), O<t<to. The reflection principle now follows from ~*
(2.11) and the fact that h(1)(x,t) can be continued into D ~ ~ by the reflection principle for the heat equation.
The uniqueness
of the continuation follows from Holmgren's uniqueness theorem. From the above analysis it is see that for the original problem the domain of dependence associated with a point in X<Sl(t) is a one dimensional line segment lying in X>Sl(t). It would be desirable to know if the restriction on the direction of the vector ~(t) can be removed. As a corollary to Theorem (2.1) we have the following version of Runge's theorem for solutions of (2.1) defined in a domain D = {(x,t): Sl(t)<x<s2(t), O<t<t } where Sl(t) and s2(t) are analO
ytic for O<t<to([3]): Corollary: Let u(x,t) e C2(D) ~ C(~) be a solution of (2.1) in
D and assume that the coefficients of (2.1) are analytic for ~<x<=, O~t@t o. Then for every c>O there exists a solution Ul(X,t) of (2.1)
in ~<x<~, O<t<to, such that ~Xlu(x,t )  Ul(X,t) l < Proof: There exists a solution Ul(X,t) s C 2 ( D ) ~ CI(~) satisfying
analytic boundary data on x = Sl(t ) and x = s2(t) such that the above inequality is valid. By reflecting Ul(X,t) repeatedly across
the arcs x = Sl(t) and x = s2(t ) it is seen that Ul(X,t) can be continued as a solution of (2.1) into the infinite strip ~<x<~, O~t~t o
65 We now turn to the case when the solution of (2.1) is known to be analytic in some neighbourhood assumption t <t<t . o o that the coefficients of the origin and make the
ytic functions of t for t <t<t for some constant to, and if (2.1) o o is reduced to the canonical form (2.4) and u(x,t) is represented in the form (2.5) we have that h(O,t) = u(O,t) Ux(0,t) and hx(O,t) = Hence by the theorem h(x,t) By
CauchyKowalewski
is analytic
in the strip =<x<=, to<t<to, yticity of the kernel E(s,x,t) Theorem 2.2:
and hence from (2.5) and the analso is u(x,t). of (2.1) be analytic in the
and let u(x,t) be any solution of (2.1) to<t<to, for some positive concontinued into the strip
~<x<~
to<t<t o.
We close this section be giving a simple derivation of Widder' result on the analytic continuation of analytic solutions of the heat equation which was used to prove Theorem 2.2 Suppose h(x,t)
is an analytic solution of the heat equation for Xo<X<Xo, to<t<t o. Then locally h(x,t) can be represented h(x,t)
= i  2hi
J Itel =
(2.14)
 2~ii
E(x,t_T)hx(O,r)d ~
JtT I = 6
66 where
oo
E(x,t) =
(2.15)
j=O (2j+l) :t j+l Since E(x,t) is an entire function of x we immediately have the result that h(x,t) is analytic in the strip ~<x<~, t <t<t .
O O X
If
we further assume that h(O,t) and h (O,t) are analytic for Itl<t and r e p r e s e n t t h e s e J u n c t i o n s by t h e i r
oo
h(O,t) =
E n= O
oo
b tn n
hx(O,t ) =
)2 Cn tn n=O
a h (x,t) n n
where hn(X,t)
a2m+l=Cm for m = O,1,2, .... Returning to the proof of Theorem 2.1 we note that from the above results a special solution of (2.4) satisfying (2.8) (for
Sl(t ) = O) can be obtained in the form u(x,t) = (~ + ~)h(x,t) where h(x,t) is given by (2.14) with h(O,t) and hx(O,t) chosen appropriately.
Ill.Parabolic
Equations
the results on the global continuation of solutions in two space variables are rather limited.
parabolic equations
in the case of analytic solutions of parabolic equations space variables with analytic coefficients whether or not such solutions
in two
67 it has been shown by Hill that the domain of dependence associated with an analytic solution of a parabolic equation in two space variables at a point on one side of a plane on which the solution vanishes is a one dimensional line segment on the other side (El6]). This suggests the possibility of establishing a reflection principle for parabolic equations in two space variables, in the case of analytic solutions. at least in
The main problem seems to be to establish the domain of regularity in the complex domain of the normal derivative of the solution evaluated along the plane on which the solution vanishes. This
suggests the problem of determining the domain of regularity in the complex domain of analytic solutions of parabolic equations, given either their domain of real analyticity or the domain of regularity of the Cauchy data in the complex domain. As will be seen,
modest results such as these can be applied to derive a Runge approximation property for parabolic equations as well as a global representation of the solution to certain noncharacteristic problems arising in the theory of heat conduction. Cauchy
Although we
shall derive these basic results in this section, we shall postpone their application to problems associated with the heat equation until Section IV. We consider analytic solutions of parabolic equations of the form + c(x,y,t)u  d(x,y,t)u t L [u~ = Uxx + u YY + a(x,y,t)u x + b(x,y,t)Uy
(3.1)
0 and for the sake of simplicity make the assumption that the coefficients of (3.1) are entire functions of their independent complex
(3.2)
z* = x  iy we can rewrite (3.1) in its complex form as
(3.3)
i i I i where A = [(a + ib), B = ~(a  ib), C = [ c, D = ~ d. The basic (or
(3.3)) is the Riemann function for (3.3) first introduced by Hill in [16]. This is defined to be the unique solution R(z,z*,t;~,~*,T)
 Vzz ,
a(AV) 3z
3(B~/) + C _ ~ r + _ ~ ~z*
(D'V') = 0 (3.4)
B(q,~*,t)d~} (3.5)
R(,z*,t;C,*,T)
= ~
exp {
along the planes z  = ~ + iQ, z* = ~* = $  in. function can be constructed by iterative methods
([3],
(under the assumption that the coefficients of (3.1) are entire) is an entire function of its six independent complex variables except for an essential singularity at t = T. equation u xx + u yy = u t the Riemann function is given by R(z z*,t;,*,T)
(3.6)
= ~
exp {
(zC) ( z *  * )
4 (tT)
} .
(3.7)
69 Now let u(x,y,t) be an analytic solution of (3.1) defined in a cylinder Dx(O,T) where D is a bounded simply connected domain. Assume that D contains the origin and that d(x,y,t) > 0 in Dx(O,T). By standard compactness arguments we can conclude that if ~
O
D,
6o > O, then u(x,y~t) is analytic in some thin neighbourhood in C 3 of the product domain box ~o,T~o3. We now use Stokes theorem
applied to u and R log r (where R is the Riemann function and r = /(x~) 2 + (yn) 2 to represent u(x,y,t) in terms of the Riemann function, where the domain of integration is the torus D
O
with ~ L =
{t:ItT [ = ~}.
Since ~
variables except for an essential singularity at t = r and H[u, R log r~ is analytic for (x,y,t) ~Do x ~ ~o' T  6 ~ , , ~ Do, ~ ~ Do,
are arbitrary):
Let u(x,y,t) be an analytic solution of (3.1) defined Then'U(z,z*,t) = u(x,y,t) is analytic in D x D* x (O,T)
where D = {z:z ~ D}, D* = {z*: z* e D}. Theorem 3.1 is analogous to the result that every regular solution of a linear ordinary differential equation with entire coefficients defined on a finite interval can be extended to an entire function
70 of its independent singularities entire. complex variable. For partial d~ffe~emtial e~fnations are
However Theorem 3.1 shows that in the case of analytic solutions in two space variables the location of the sinin a simple way by the
inition of the solution in the real domain. in two independent variables Bergman
in terms of the Riemann function and the Goursat data on the characteristic hyperplanes z = O and z* = O in a manner almost identical to the
use of the classical Riemann function to solve the Goursat problem for hyperbolic equations in two independent variables. Since this Goursat
data is defined in a product domain we can apply Runge's theorem for several complex variables sets by polynomials to approximate this data on c o m p a c t s u b 
and hence, since the Riemann function is entire on compact subsets by an > O every
solution of a parabolic
defined in a cylindrical
imated by a solution having analytic boundary data, and hence from the Introduction is an analytic solution of the parabolic equation,
we can now deduce the following version of Runge's theorem ([3]): Corollary: Let u(x,y,t) be a classical > O in D x (O,T). solution of (3.1) in D x (O,T)
where d(x,y,t)
and positive constants ~ and e there exists an entire solution Ul(X,y,t) of (.3.1) such that
71
max o
lu(x,y,t) Ul(X,y,t)I
< e .
the
Cauchy pro
blems for parabolic equations with data prescribed along an analytic surface. Such problems arise if inverse methods are used to study (c.f. Section
IV) and a local solution can always be found by appealing to the CauchyKowalewskl theorem. However in addition to being impractical for com
putational purposes such an approach does not provide us with the required global solution to the Cauchy problem under investigation, and hence we are lead to the problem of the analytic continuation solutions to noncharacteristic We shall accomplish of
this through the use of the Riemann function in integration techniques and the calculus of
residues fn the space of several complex variables. We assume that u(x,y,t) is an analytic solution of (3.1) in a analytic surface Let
domain containing a portion of the noncharacteristic S along which u(x,y,t) the intersection one dimensional F(x,y,t) = O. assumes prescribed
of the plane t = constant with the surface S be a curve ~(t). Suppose ~(t) is described by the equation
(3.10)
where z and z* are defined 5y (3.2) and this is the equation for ~(t) in (z,z*) space. We now choose C(t) to be an analytic curve lying on not on S let G(t) be a
72 cell whose boundary consists of C(t) and line segments lying on the characteristic planes z = ~ = ~+i~ and z* = ~ = tin respectively which join the point (~,~) to C(t) at the points P and Q respectively. If we now use Stokes theorem to integrate RL[I~r]~rM[R] (where R is
the Riemann function and L and M are defined by (3.3) and (3.4) respectively) over the torus {(z,z*,t):(z,z*) ~r(~,~,T) = ~. . ~ .. 4~i J e G(t),ItT I = 5} we have
[ R ( P , t ) ~ (P,t) + R(Q,t)7F(Q,t)]dt
It~1 =~
2~i
(3.II)
where we have suppressed the dependence of the Riemann function on the point (~,~,T) and an expression of the form 7jr(P,t) is a function of three independent variables, i.e.7_~(P,t) =~r(~l,~2,t) where ( 1, 2) are the Cartesian coordinates of the point P in C 2. For (~, ~,~) sufficiently near the initial surface S and for sufficiently small, (3.11) gives an integral representation of the
solution to the noncharacteristic Cauchy problem for (3.1) with analytic data prescribed on S. Equation (3.11) can now he used to
obtain a global solution by deforming the region of integration, provided a knowledge of the domain of regularity of the Cauchy data and analytic function F(x,y,t) is known. In particular such a pro
cedure yields results on the analytic continuation of solutions to parabolic equations along characteristic planes in terms of the domain of regularity of the Cauchy data and the domain of regularity of the function describing the noncharacteristic surface along which
73 the Cauchy data is prescribed. For example suppose for each fixed
t, t e [O,TJ, z = $(~,t) maps the unit disc conformally onto a domain Dt and let the analytic surface S be described by im ~l(z,t) = 0 (3.12)
Assume that (z,t) and l(z,t) depend analytically on the parametec t. Then setting ~(~,t) = (~,t) we have that the equation Hence if we assume
that the Cauchy data is analytic in Dt for each t we have from (3.11) the following result (c.f.~]): Theorem 3.2: Let u(x,y,t) be an analytic solution of (3.1) which
assumes analytic Cauchy data on the surface (3.12) where @(z,t) conformally maps the unit disc onto the domain Dt. If the Cauchy
data for u(x,y,t) is analytic in Dt then, for each fixed t, 7J(z,z*,t) = u(x,y,t) is an analytic function of z and z* in Ot x Dr* where O t" = {ze:z* e Ot). Theorem 3.2 is a generalization of the corresponding result obtained by Henrici for elliptic equations in two independent variables ([lhJ). For partial progress on extending the results
of this section to parabolic equations in three space variables we refer the reader to [23S.
IV. Applications to Problems in Heat Conduction. In this last section we shall apply the results of the previous two sections to derive an explicit analytic representation of the solution to the inverse Stefan problem and to construct some complete families of solutions to the heat equation which are suitable for constructing approximate solutions to the standard initialboundary value problems arising in the theory of heat
74
conduction (c.f.[3]).
This
is a free boundary problem for the heat equation and we are interested in the inverse problem where the free boundary is assumed to be known a priori. Such an approach allows one to construct a
variety of special solutions from which a qualitative idea can be obtained concerning the shape of the free boundary as a function of the initialboundary data. In certain physical situations, e.g.,
the growing of crystals~ the inverse problem is in fact the actual problem that needs to be solved. The simplest example of the type
of problem we have in mind is the single phase Stefan problem for the heat equation in one space variable, which can be mathematically formulated as follows: u to find u(x,t) and s(t) such that xx = ut; O<x<s(t), t>O t>O t>O t>O The function u(x,t)
(2.z)
is the temperature in the water component of a onedimensional icewater system, s(t) is the interface between the ice and water, and (t) is assumed to be given. The inverse Stefan problem assumes
that s(t) is known and asks for the solution u(x,t) and in particular the function (t) = u(O,t), i.e., how must one heat the water in order to melt the ice along a prescribed curve? If we
assume that s(t) is analytic the inverse Stefan problem associated with (h.i) can easily be solved using the results of Section II. Indeed if in (2.1h) we place the cycle It~l = 6 on the two dimensional manifold x = s(t) in the space of two complex variables, and note that since u(s(t),t) = 0 the first integral in (2.1h)
75 vanishes, we are lead to the following solution of the inverse Stefan problem: u(x,t) = 12wilt~l=6 E(xs(T),tx)s(T)dT
(4,2)
Computing the residue in (4.2) gives u(x,t) = ~ I n=l ~   S t Sn n [xs(t)] 2n , (4.3) The
idea of the inverse approach to the Stefan problem (4.1) is to now substitute various values of s(t) into (4.3) and compute (t)=u(O,t) for each such s(t). For example setting s(t) = /t gives n! (t): Z ~ K , n=l = constant , (4.4)
a result corresponding to Stefan's original solution. We now consider the Stefan problem in two space variables. equations corresponding to (4.i) are now Uxx + uyy = ut; uI so x (x,y,t)<O The
= y(x,y,t)
(4.5)
uI
~=0
S
= o
Su I =0
St =O
where u(x,y,t) is the temperature in the water, (x,y,t) = O is the interphase boundary, D is a region originally filled with ice, y(x,y,t) ~ 0 is the temperature applied to the boundary SD of D, is the normal with respect to the space variables that points into
76 the water region $(x,y,t) < O, and V denotes the gradient with respect to the space variables. In this case from the analysis of Section !II
we see that in order to solve the inverse Stefan problem it is necessary to restrict the function #(x,y,t) to be more than just analytic, and we assume (x,y,t) is given by an equation of the form (3.12) where Dt D D for t ~ [O,T]. In this case we have from (3.11) and
7 J (~, ~,T) : ~
t~
exp [(z~)(z*~q
[~h(tiT') ~
IV$1 St
~ ,dzidt
I ~I:a
c()
(~.6)
where C(t) is a curve lying on the surface $l(z,t) = l(z*,t) with endpoints on the characteristic hyperplanes z =~ and z* = ~. Computing the residue in (h.6) now gives
7j(~,L+)
: +
)a
~n ~Tn
{j
(z~)n(z*~) n
3

n:0 hn(n!
twl
Id+l} (~.7)
c(+)
Equation (h.7) gives the generalization to two dimensions of the series solution (h.3) for the one dimensional inverse Stefan problem. We note that the integral in (h.7) is pure imaginary. We now turn our attention to the construction of complete families of solutions for the heat equation. By completeness we
shall mean completeness on compact subsets of a given domain D with respect to the maximum norm. We first consider a solution
h(x.t) of the one dimensional heat equation (2.6) defined in a domain D of the form D = {(x.t):sl(t)<x<s2(t) , 0<t<t O} where sl(t ) and s2(t) are continuous functions. without loss of generality that D o If % c D then we can assume
hence from the Corollary to Theorem 2.1 we have that for every
77 E ~ 0 there exists a rectangle R P heat equation in R such that and a solution hl(X,t) of the
(~.8)
nomials h (x,t) defined by (1.7) are complete for solutions of the heat equation defined in a rectangle, we can now conclude that the heat polynomials are also complete for solutions of the heat
equation defined in a domain of the form described above. We next want to decide under what conditions on the separation constants k n do the solutions + hn(x,t) = exp (+_ i~nX  ~2t)n form a complete family of solutions From the above results (4.9)
nomial can be approximated by a linear combination of the functions (4.9), subject to certain restrictions on the constants ~ . n From
the representation (2.1h) we see that it suffices to show that the 2 set {e ~nt} is complete for analytic functions defined in an ellipse containing [O,to]. The type of restriction necessary is indicated
(B@ ,p.219):
Theorem h.l: the limit d = lim n'~ n ~ ~n > 0 If {~n } is a sequence of complex numbers for which
exists, the syste~ {e ~nz} is complete in the space of analytic functions defined in anY region G for which every straight line parallel to the imaginary axis cuts out a segment of length less
78 than 2wd, and the system is not complete in any region which contains a segment of length 2~d parallel to the imaginary axis. From Theorem 4.1 we now see that the set (4.9) is complete for solutions of the heat equation defined in a domain D of the form described above provided lim n~ n ~2
n
> 0 .
(4.10)
Theorem 4.2:
for solutions of the heat equation defined in D: [n/~ Z k=O xn2ktk ........ (n2k)!k!
I) h n(x,t) = [~]'.
; n = 0,1,2 .....
We now conclude this section by using the results of Section III to derive a result analogous to Theorem h.2 for the heat equation in two space variables defined in a cylindrical D is a bounded simply connected domain. domain D x [O,T] where
to Theorem 3.i we see that it suffices to obtain a set of solutions to (3.6) assuming data on the complex hyperplanes z = 0 and z* = 0
that is complete for product domains in the space of two complex variables. It is easily verified that one such complete set is given by + U~,m(r,e,t) +ine = em r2k+n tmk Z k=O 4kk~(mk)!(n+k)! (h.ll)
79 zntTM
+ (z , O , t ) . . . . . . n,m m'n'
(h.12)
V n,m ( O ' z * ' t ) : z*ntm m'n'
It follows from the uniqueness theorem for Cauchy's problem for the heat equation that another complete set for (3.6) defined in D x (O,T) is given by hn,m(X,y,t) = hn(x,t) hm(Y,t) where hn(x,t) is the heat polynomial defined by (1.7). (h.13) On the
other hand if we separate variables for (3.6) in polar coordinates we find that
+
h, m (r,e,t) = Jm(Anr)exp(+im8  X~t) n
m
(h.lh)
where J (z) denotes Bessel's function, is a solution of the heat equation satisfying the complex Goursat data z H~n+'m(Z'O't) exp( ~ t) 2m m:
(h.15)
Hn+m(Z,z*,t)
= hn ~ m ( r , B , t ) +
and hence from Theorem h.l we can conclude that (h.lh) is another complete set of solutions for the heat equation in D x (O,T) provided (~.10) is valid. Theorem h.3: plane. Let D be a bounded simply connected domain in the
heat equation defined in D x (0,T): l) hn,m(X,y,t) = hn(x,t)hm(Y,t) ; n,m = 0,1,2,... 2) h+,m(X,y,t) = Jm(Xn r) exp(+_ ires  X2t)n;n'm= O,i ....limn~<o >Vn O. Xn
80 Theorem h.3 can also be extended to higher dimensions ([6~). For other methods of proving Runge's theorem for the heat equation see [18~ and [22].
References 1. S. Bergman, Integral Operators in the Theor~ofLinear Differential Equations, SpringerVerlag, Berlin, 1969. 2. D. Colton, Partial Differential Equations in the Complex Domain, Pitman Publishing, London, 1976. D. Colton, Solution of Bound ar~ Value Problems by the Method of Integral Operatgrs , Pitman Publishing, London, 1976. D. Colton, A reflection principle for solutions to the Helmholtz equation and an application to the inverse scattering problem, Glas~ow Math ,. J. 18(1977), 125130. D. Colton, On reflection principles for parabolic equations in one space variable, Proc. Edin. Math. Soc., to appear. D. Colton and W. Watzlawek, Complete families of solutions to the heat equation and generalized heat equation in J. Diff. Eqns~ 25(1977), 96107. V. De Alfaro and T. Regge, Potential Scattering, NorthHolland Publishing Company, Amsterdam, 1965. A. Erdelyl, Singularities of generalized axially symmetric potentials, Comm.: Pure Appl. Math. 9(1956), hO3hlh. " V. Filippenko, On the reflection of harmonic functions and of solutions of the wave equation, Pacific J. Math. lh (196h), 883893.
J
Partial
R n,
81 i0. A. Friedman, Part%al Differential Equations, Holt, Rinehart and Winston, New York, 1969. P. Garabedian, An example of axially symmetric flow with a free surface, in Studies in Mathematics and Mechanics Presented to Richard yon Mises, Academic Press, New York, 195h, 149159. 12. P. Garabedian, Partial differential equations with more than two independent variables in the complex domain, J. Math. Mech. 9 (1960), 2hi271. 13. R . P . Gilbert, Function Theoretic Methods in Partial Differential Equations, Academic Press, New York, 1969. P. Henrici, A survey of I. N. Vekua's theory of elliptic partial differential equations with analytic coefficients, Z.An~ew.Mat h. Physics 8(1957), 169203. C. D. Hill, Parabolic equations in one space variable and the noncharacteristic Cauchy problem, Comm. Pure Appl. Math. 20 (1967), 619633.
ii.
15.
16.
C. D. Hill, A method for the construction of reflection laws for a parabolic equation, Trans. Amer. Math. Soc. 133 (1968), 357372.
17.
F. John, Continuation and reflection of solutions of partial differential equations, Bull. Amer. Math. Soc. 63 (1957), 3273hh.
18.
B. F. Jones, Jr., An approximation theorem of Runge type for the heat equation, Proc. Amer. Math. Soc. 52 (1975), 289292. B. Levin, Distribution of Zeros of Entire Functions, American Mathematical Society, Providence, 1964.
19.
82 20. H. Lewy, On the reflection laws of second order differential equations in two independent variables, Bull. Amer. Math. Soc. 65 (1959), 3758. 21. H. Levy, On the extension of harmonic functions in three variables, J. Math. Mech.lh (1965), 925927.
22.
teoremi di unicit'a e
teoremi di completezza connessi col metodo di integrazione di M. Picone, Rend. Sem. Mat. Univ. Padova 21 (1952), I, 99123, II, 136170.
23.
M. Stecher, Integral operators and the noncharacteristic Cauchy problem for parabolic equations, SIAM J. Math. Anal. 5 (1975), 796811.
I. N. Vekua, New Methods for Solvin~ Elliptic Equat%ons, John Wiley, New York, 1967.
25.
D. Widder, Th9 Heat Equation, Academic Press, New York, 1975. D. Widder, Analytic solutions of the heat equation, Duke Math. J. 29 (1962), h97503.
26.
W N Everitt
Introduction
number of different ways which we review here briefly before discussing the connection with differential operators. For convenience
let (I,) be the open interval of the real line, and let N o = {0,1,2,. .... } be the set of nonnegative integers. All the definitions given below are interconnected, as may be
seen in the standard accounts of the Legendre functions given in [2] by Erdglyi et al, and [9] by Whittaker and Watson. (i) The Legendre differential equation For our purposes this is written in the form (here ' ~ d/dx) (x ~ (l,l)) (I.I)
which is derived from the Laplace or the wave equation, considered in polar coordinates. This equation has singular points at l where This equation has a nontrivial
solution which is bounded on (I ,I) if and only if n No; the corresponding solution is P (), i.e. the Legendre polynomial of order n. n See [2, chapter III] and [9 sections 15.13 and 15.14] (ii) The Poisson generating function This definition takes the form
I /(I  2xh + h 2)
= ~ Pn(x)hn n=
(x (I ,I))
Pn(x) 
1 2 n n. v
dn
dx n
(x 2
1) n
(x
(1,1),
~ No);
84
(iv)
The GramSchmidt
ortho$onalization
The Legendre polynomials may be defined by the GramSchmidt process applied to the set {xn : x ~ (l,l), n E N }
o
in the integrablesquare
innerproduct
space L2(1,I).
See Akhiezer
and Glazman [1, section 8], [2, sections [9, section ll,6]. This definition
1
where ~ is the Kronecker delta function. mn Whichever definition is adopted it is important subsequently to prove that the set of Legendre polynomials is complete, equivalently closed, in L2(1,1); and [2, section I0.2]. {P (.):n e N o }
n
is known the completeness may be obtained by classical means, as for the completeness of the trigonometrical functions, using the Weierstrass
Our interest in this paper is to discuss the definition and completeness of the Legendre polynomials in L2(I,I) from the
%wy on
(a,b)
(1.3)
coefficients
on the interval
(a,b),
then this is a socalled risht definite problem and is studied in the integrablesquare f f or which space ~ ( a , b ) , i.e. the collection of those functions
(1.4)
If in (1.3) it should happen that, whether w is of one sign or not, both p and q are nonnegative then the problem is called
85
leftdefinite and is studied in the space of those functions f for which b f {p(x) If'(x)I2 + q(x) If(x)12}dx < ~
a
all solutions of (1.3) are in the spaces in the neighbourhood the classification
definite case see Naimark [4, section 18.1] and Titchmarsh [8, sections 2.1 and 2.19]; and in the leftdefinite [6, sections I . 4 and 5]. For the purposes of the study of the Legendre equation in this paper we write the equation in the form ((I  x2)y'(x)) ' + y(x) = % y(x) (x ~(l,l)), (1.1) case see Pleijel
(1.6)
and
so that in comparison with the standard form (3.3) we have a = I, b =  2 p(x) = I  x q(x) = w(x) =  (x c(1,3)). in the
Thus we may study the Legendre equation as rightdefinite space L2(I,I), and leftdefinite in the space
I
I I
which, for convenience in this paper, we denote by H2(;,I). The original study of Legendre's differential equation in the rightdefinite 4.3 to 4.7]. limitcircle case in L2(l,l) is due to Titchmarsh; see [8, sections (1.6) is
It should he noted that the Legendre equation at both endpoints ~I in the rightdefinite
case; for
details see [8, section 4.5] and [I, II, appendix II, section 9, II]. Here it is emphasized that the analysis of Titchmarsh is essentially 'classical' with no reference to operator theoretic concepts. This
work was followed by the studies of Naimark [4] and Glazman as given in [3, appendix II, section 9]; in particular Glazman characterized elements in the domain of the differential Legendre polynomials. The first study of Legendre's differential equation in the leftdefinite case is due to Pleijel; see [5] and [6] in which may be the
86
found references to earlier results of Pleijel and the work of his school at Uppsala. In particular we owe to Pleijel the observation
that the Legendre equation (1.6) is limitpoint at both endpoints I in the leftdefinite case; see [6, page 398]. Our purpose in this paper is to study the right and leftdefinite problems for the Legendre differential equation with the methods of Titchmarsh [8] in mind. We link the Titchmarsh method
with operatortheoretic results in the Hilbert function spaces L2(,]) and H2(I,I). The paper is in six sections; after this introduction the second section considers essential properties of the Legendre differential equation; the third and fourth sections are given over to r#mgrks. a study of the right and leftdefinite cases; five and six to c e r t a i n / Notations ACIo c R  real field; C  complex field; L  Lebesgue integration;
'(f D)' is to be read 'the set of all f D'; NO  the set of all nonnegative integers. 2. The Legendre differential equation In this section we consider certain essential properties of Legendre's differential equation required for consideration of the rightdefinite and leftdefinite cases. As before we write the equation in the form ~I  x 2 ) y ' ( x ~ ' + y(x) = % y(x) For convenience let
p(x) = J  x2 (x ~ (l,I)). (2.1)
(x (I,])).
(1.6)
The standard form of the Legendre equation is given in (1.I) but for the purpose of considering both right and leftdefinite problems the form (1.6) is to be preferred. A detailed discussion of the classical
properties of the Legendre equation (I.I) is to be found in Erd~lyi et al [2, volume I, chapter III]. The account of the Legendre equation in Titchmarsh [8, chapter IV] is based on the Liouville normal form of (1.6) i.e.
~ 2
y"(x)  sec x.
y(x) = ~ y(x)
(x (~,n~ .
However, as is evident from the results in [8, section 4.5], this differential equation does not of itself enjoy the property of having
87
polynomials
solutions.
of Titchmarsh
to
when
(2.2)
two solutions
the analysis
of (1.6) above may be determined by I COS X cos st dt Y(X, ~) = [ J I (cos t  x) I/2 COS x
(x (l,l)
~ C) (2.3)
Z(x, ~) =
functions
from ~ to O, and sinlx to increase from In to ~ as, in both cases, x increases froml to I.
To show that the Y and Z, as defined by (2.3), are solutions of (1.6) we follow [8, section 4.5] and write the integrals
J G (cos z  x )  / 2
cos sg d z
Z ( x , ~) = ~ J F (cos z + x) I/2
where now the sign convention
(2.5)
for Y the integrand is made singlevalued by cutting the zplane from 1 I  cos x to cos x; similarly for Z the cut is from  (~ + sinlx) to ( ~ + sinlx). Here the contours G and F can be
taken as circles with centre the origin of the zplane and of radius r and p respectively, where coslx < r < ~ and ~ + sinlx < p <~.
88
the method
of [8, section
the integral
to show that Y and Z are solutions and all % ~ C . (2.4) and (2.5)
of the differential
equation
of the inverse
functions
(x ~ (1,1)
~ ~ c).
Y(O,%)
= Z(O,%)
(2.6)
Y'(O,%)
25/2~;/2 =  r( + s) r(  s)
= Z'(O,%).
(2.7)
Y and Z
= O, i.e.
of x and has
the value,
z(o,~) = 2z(o,~)
z'(o,~)
(2.9)
= 8~ cos ~s on following the analysis in [8, section forms 4.5]. Y and Z in the we give
The asymptotic neighbourhood some details of the singular for the point
of the solutions
endpoints
 and there
for l.
Y we find,
Y'(x,X)
~ (%  )~//2
representation
the contour
and we obtain
lim x I
(~
C).
80
Notwithstanding
and on G except for a simple pole at the origin. residues yields lim x+lSimilarly lim x + lY'(x,~) = ~ [ JG ( c o s z COS SX
Y(x,%)
= 7/2
(~ e C).
dz = (~)~//2
1) 3 / 2
also valid for all % e C. For the solution Z we find Z(x,%) ~ 2/2. cos ~s. In ('/(lx) (x + 1)
(2.11)
cos ~s.
(I  x) I
(x + I)
both valid for all % c C except for the set of points {(n + i)2 : n = O, I, 2 . . . . }. The proof of the first of these results
follows from the analysis given in [8, section 4.5]; the second result then follows from the constant value of the Wronskian given by (2.8) and (2.9). The asymptotic above results forms of Y and Z at  follow from the Y(x,%) = Z(x,%) (x e (1,1)). and the linear of Y and Z as
of the solutions Y and Z, except for certain exceptional (1.6) has a nontrivial solution
values of %, that the Legendre equation which is bounded on the internal set of points
{(n + )2 : n = O, I, 2 . . . . }.
From Y and Z we now form solutions O and ~ of the Legendre equation (1.6) which satisfy the following
= 1 9'(0,~) = 0 (p(O,~) = 0
initial conditions
= . (2.12)
0(0,~.)
~'(O,~k)
In fact we have
O(x,~,) = Y ( x , % ) + Z ( x , k ) 2Y(O,~) '~ (x,%) = Y ( x , ~ )  Z(x,~ .) 2Y' (0,%) (x E (  1 , 1 ) ) ( 2 . 1 3 )
for all ~ ~ C (but with care needed at the set {(n + )2 : n = 0, I,2,...}).
90 From the general properties of the differential equation (1.6) we know that the Wronskian p(@~'  0'~) is constant
on (I,I) and so
(1  x 2 ) ( e ( x , ~ ) ~ ' ( x , k )  @'(x,~)~(x,~) = 1 (x e (  l , l ) , k ~ C)
The asymptotic forms of 0 and ~ follow from the earlier results for Y and Z given by (2,10) and (2.11). For
x ~
]
1 x
(2.15)
(2.16)
I/2
~'(x,l) ~  2F(% +' ~s')'F(% s) "
(2.17)
all these results are valid at the set {(n + 3) 2 : n = 0,1,2 .... } of the kplane. With these results established it is now possible to look at the classification in the rightdefinite, section ; above. In the rishtdefinite case; from the asymptotic results of the Legendre differential equation cases, as introduced in (1.6)
and leftdefinite
above for 0 and ~ it is clear that although these solutions are unbounded, in general, in the neighbourhood of the endpoints 1, both @ and
are in L2(O,I) and L2(I,0) for all % e C; thus the equation is limitcircle at both ! and I in this case. (This result may also be (x e (  1 . 1 ~
: both
established independently
[8, section 2.9] the equation is limitcircle at 1 and I.) In the leftdefinite case; here the spaces concerned the asymptotic results as given above, show
are, see section ! above, H2(O,I) and H2(l,O); for the solutions 0 and ~ and their derivatives,
that for all nonreal ~ e C neither @ nor ~ is in H2(0,I) or H2(;,O) and this implies that the equation is limitpoint case. (In the leftdefinite at 1 and I in this
9~
at the nature of solutions at real values of %, in order to determine the classification of the equation; we do not discuss this point here but see the works of Pleijel [5] and [6], and references therein.) 3. The rishtdefinite case. In this section we consider the rightdefinite This section
is dependent in part on the original analysis of Titchmarsh in [7] and later in [8, chapter II, section 4.4]. Consider the Legendre differential equation in the form (1.6) on the interval (1,1) i.e.  ((lx2)y'(x)) ' + y(x) = %y(x) Cx (1,1)).
For solutions of this equation in the neighbourhood of the singular point l, the general theory in [8, chapter 2] gives the existence of the TitchmarshWeyl mcoefficient; the analytic function m(') : C C
and determines a p~rticular solution ~ of the equation in the form ~(x,%) = e(x,%) + m(%)~(x,%) x (;,I) % C R , (3.;)
As the differential
equation is limitcircle at both I, the mcoefficient is not unique and, in order to determine the differential operator associated with the Legendre polynomials, we have to make a suitable choice from the family of mcoefficients belonging to the endpoint I. We do this by
following the limit process which determines m(') from the 1functions; for the general theory see [8, section 2.;]. With the solutions 0 and ~ determined from (2.;2) and with % ~ C R, let ~ X be a solution of (1.6), here X (x (1,1) . (0,1), given by
The function i is chosen so that ~X satisfies the following boundary condition at X @X(E,%) cos B + P(X)~x'(X,%)sin B = 0 for some B (~,~] ; thus, for all ~ C~R, I(%,X,B) =

Now let X + ; and choose ~ as a function of X so that 1 tends to a limit m(.), where m() : C + C and is regular on c~R. In this the
92 Legendre case we can see how this is done explicitly = } + p(X)Ol(X,l ) tan ~(X).{in((lX)l)} I tan B(X).{In((IX) l)}I we have (3.2) by writing I in
_ O(X~I){In((IX);)
tan B(X).{In((IX) I) } + tan y as X the limits of all the remaining formulae (2.14 to 17) .
In our case, guided by [8, section 4.5], we (O,l)); it then follows that
where we recall,
see (2.2), that s = /A. (3.3) we then find see (3.1), (x ~ (I,]) (I e CNR) I e ckR) that the resulting solution 4(.,%)
and 4'(',I)
(1 c CxR).
A similar
analysis
endpoint
l;
X ( X , 1 ) = O(x,%) + n ( t ) ( x , t )
such that X(.,I) has the properties (i) (ii) X(X,1) = Z(x,1)/Z(0,%) X(',I) ~ L2(I,I) X(X,l) (x E (l,l) I ~ CxR)
(I CNR) s)
= ~I /2F(% + s ) r ( % 
83
(iv) (v)
 )1`(% + ~ s ) r ( % 
s)/(2~) 1
a r e unbounded i n t h e n e i g h b o u r h o o d of
(X C\R). The Green's function for this choice of m() and n(.) is given by G(x, ; X) ffi
=
_ ~ ! x , ~) ~ (~, ~) {p W(~,~)}(%)
X(~) ~ (x,~)
(1
<x
<g
< 1)
(1
<g
<x
<1)
(~(x,X)'(x,k)
 m(~) (~
 ~'(x,~)(x,l)
C~R).
(x (I,I)
From the general theory of differential equations of the form (1.6) with two singular endpolnts, that the eigenvalues it is known, see F8, section 2.18], choice of In this
and X above, are given by the zeros and poles of n(')  m(). case, from (3.4), n(%)  m(X) =  2m(%)
= 8 I'(% + s ) F ( % s)
(3.5)
r(~ + Is)r(
and from this a calculation shows (recall s = ~; eigenvalues are given by %n = (n + )2 Anticipating (n No).
the set of eigenvalues given by (3.6). Following the analysis in [8, section 4.53 it may be shown that the eigenfunetlons P~(T) are given by ~n(X) = (n + ~)I/2en(x ) (x ~ (I,I) n No) (3.7) see [2, sections 3.6.2 ~ {~n(. ) : n ~ N o } corresponding to the eigenvalues
where {P () ; n ~ N } are the Legendre polynomials, n o and 10.6] and [9, section 15.1].
Here we leave the classical study of the differential equation of Legendre and turn to the study of the associated differential
94
operator in this, the rightdefinite, case. Let the symmetric differential expression M be defined by M[f](x) = ((I  x2)f'(x)) ' + f(x) (x ~ (1,1)) For any f and g
for any f : (I,I) ~ C with both f and f' E ACloc(1,1). with these properties, integration by parts shows that
B
I J {~ M[f]  fM[g]} = [fg](')l~ [~,B] c (l,l), where  f'(x)g(x)) (x c (l,l)). (3.9) (3.8)
[fg](x) = p(x)(f(x)g'(x)
We define a differential operator T, later shown to be selfadjoint in the Hilbert function space L2(l,l), as follows:
secondly define the domain D(T) c A by f e D(T) if (iv) f A and for some I c C\P ~ (T) (a) lim
x+ 1
If(x) ~ (x,l)] = 0
(b)
lim
xI
If(x) X(x,I)] = O ;
Note that the limits in (iv)(a) and (b) exist and are finite in view of the Green's formula (3.8), and since f, ~ and X all satisfy conditions (i) and (iii) for A. The genesis of (iv)(a) and (b) as the correct general
form of boundary condition at a singular endpoint may be seen in [8, section 2.7], and receives its full development in Naimark [4, section 18]. conditions determined, Note also that ~ and X themselves satisfy the boundary are
and (b); these results follow essentially from the important result given in [8, lemma 2.3].
95
Similar analysis also shows that for all f, g E D(T) (but not for all f, g e A ) lim If g](x) = O
xl
lim If g](x) = 0
x'+I
(3.0)
and then from Green's formula (3.8) it follows that (Tf,g) = (f,Tg) (f,g E D(T)) (3.11)
where (,.) is the innerproduc~ in L2(I,). If Co(l,l) represents all infinitely differentiable functions with compact support in (],l) then clearly Co(l,;) c D(T); hence D(T) is dense in L2(,I). that T is symmetric in L2(],I). If also we define ~ : (1,1) x C\R x L2(I,]) C by, see [8, section 2.6] (x,l;f) = From this result and (3.11) it follows
I'
I
G(x,~;l)f(~)d~
(3.la)
then ~ , % ; f )
from this result it may be shown that (T i il)D(T) = L2(I,;). Thus T is a selfadjoint (unbounded) differential operator in the space L2(l,l). We now give a number of different but equivalent descriptions of the elements of the domain D(T), and, in particular, give a number of alternative forms of the boundary conditions (iv)(a) and (b) which reduce A to D(T). For an alternative account of these boundary conditions see
[, appendix II, section 9, example II]. We first show that if f D(T) then lim f exist and are both l We have from [8, lem~na 2.9], but with our sign convention, for
finite.
it is essential for this result to hold that f satisfies the boundary condition (iv)(a) and (b). Next we prove the general result that if g ~ L2(1,I) and E C\Po(T) then lim ~ (.,~;g) both exist and are finite; for I
96
fx
I
X(~,X)f(~)d~ + X(X,%)
11
x
@(~,X)f(~)d~
IX (x,X) x*(~'X)f
d~.xlg(~)l 2d
= O(ln ((lx)l).
= o(1) (xl)
on using the results in (2.10 and ll). lim Ix ~2 x~l ~(x,X) _iX(~,X)g(~)d~ =Y~,X) which is finite.
(lx) 1/2)
Also
I~IX(~,X)g
(~)d~
Thus it now follows from (3.12) that if f D(T) then lira f both exist and are finite; if f(l) is defined by these limits then I f c C[I,I] for all f ~ D(T). Now suppose that f e A and lim f exist and are both finite; then we state that lim pf' = O. + To see this we can suppose, without loss
of generality, that f is realvalued on (I,I); we have M[f] E L2(O,I), i.e. M[f] ~ L(O,I) and (pf')' ~ L(O,I); hence for some real k lim
Pf' = (pf')(O) +
I1
(pf')' = k;
(3.13)
O now if k ~ 0 then we can take k > 0 and obtain f(x) ~ f(O) + ~k I~pI for x close to I, i.e. lim f = = which is a contradiction; hence k = 0 1 and lim lim 1 pf' = O; similarly I pf' = O. lim Now suppose f e ~ and I Pf' = O; then If(x) l ~ If(o) l + 0 If(o) l + K in ( ( I  x) I) (x ( [ o , I ) ) If'l = If(o) l + Ipf'l
for some positive real number K; this last result, and again using lim 1 pf' = O, together with the known properties of ~, proves that lira if ~ (. ,l)] = O for any % ( C\Po(T); similarly lim [f X (,l)] = O; I
hence f E D(T).
97
If now f E D(T) then, from the results overleaf, lira pf,~ = O; hence from + {plf'
X
+ Ill 2} = pf,.?
.f
(3.14)
L2 t_i,I~; ~
then M[f] e L(o,I) and as in (3.13) lim pf, = k (say); if k ~ 0 then + for x close to I it follows that, for some k
o
> 0,
iXp ]f,[2
as above, f ~ D(T).
zk O
iP x.
Taking all these results together it follows that the domain D(T) can be described in any one of the follow%ng five equivalent forms f ~ D(T) if f A and either (~) lim [f~] = lim [fx] = 0 1 I or or or or (8) (y) lim +I f exist and are finite lim _+ pf' = 0
where in (Z) the notation 1 is used to represent the function taking the value 1 on (1,I). We can prove a little more; if f c D(T) then from (3.14) above we see that {plf' I + Ill2} = 1 M[f]'f (f c D(T))
so that M satisfies the socalled Dirichlet formula on D(T) hut not, as may be readily shown, on the maximal linear manifold A. formula we see that the selfadjoint operator T satisfies (rf,f) > (f,f) (f E D(T)) (3.15) This is a special From the Dirichlet
98
bounded below; in fact the first eigenvalue %o of T is . We comment on the spectrum of the selfadjoint operator T;
this consists of the set Po(T) = {%n = (n + )2; n E No}, see (3.6), each point of which is a simple eigenvalue with eigenfunetions Legendre polynomials and, in particular, the
{Pn(') : n No}; clearly Pn(') D(T) (n No) satisfies the boundary conditions (iv)(a) and (b). of the
solutions ~(,~) and X(',~), given earlier in this section, that no solution of Legendre's differential equation (1.6), with % = ~, can be
found which satisfies the boundary conditions at both singular endpoints +l. Indeed at all points D e R\P(T) it may be shown that (T  ~ I)D(T) = L2(],l), on using the result (3. ;2); this shows that
is in the resolvent set of T; see if, section 43]. The general spectral theory of selfadjoint see [ 1, chapter VII now yields the completeness polynomials in L2(l,l), operators,
if we define the
operator S : Co(l,;) L2(l,l) by Sf = M[f] (f Co(l,)) and satisfies the inequality see (3.15). The general theory of semi
(Sf,f)
> ~(f.f)
(f E Co(l,l)),
bounded sy~netric operators then applies, see [I, section 85], and the operator T then appears as the uniquely determined Friedrichs extension of S; this relates to the form (6) of the equivalent boundary conditions, i.e. a finite Dirichlet condition. 4. The leftdefinite case. We again consider the Legendre differential
equation in the form (1.6) M[y](x) = ((I  x2)y'(x)) ' + ~y(x) = %y(x) As in section I above we define H2(I,) = (x c (l,l)). (1.6)
99
We noted in section 2 above that the differential expression M is limitpoint in H2(I,I) To obtain a selfadjoint at both the singular endpoints I. operator S, say, in H2(I,I),
as generated by M and playing the same r~le as the operator T in section 3, we follow the method used in Everitt [3], using also, in part, the work of Atkinson, Everitt and Ong in [ IO], There is a theory of the mcoefficient for leftdefinite of the We again use
problems which reflects some, but not all, of the properties TitchmarshWeyl mcoefficient in the rightdefinite theory.
the solutions O and ~ of (1.6) introduced in (2.13). Since the differential expression M is limitpoint in H2(O,I) at the endpoint H2(O,I) for any ~ e C\R. I, neither solution @(',~) or ~(.,l) is in However there exists a unique coefficient m(') case) which is
(we use the tilda notation to distinguish the leftdefinite analytic (regular, holomorphic) = @ + m ~ ~ H2(o,I). independent
Now for )~ c C\R there is only one linearly (1.6) which lies in H2(o,I); from
the asymptotic results (2.10 and 11), for the solutions Y(',%) and Z(',%), it is clear that this solution in H2(O,]) must be Y(',%). ~(.,~.) = 0(,%)
+ m(%)~(.,%) = k(~)Y(,%)
Thus
If we differentiate
Similarly at the endpoint I the solution in H2(I,o) is Z(.,%) and, writing ~ = @ + n q~,
~(~)
rightdefinite
z'(o,~)/z(o,~)
 ~(~).
(4.3)
Note that m = m of (3.3), and similarly n = n, of the case but note that m is unique whilst we had to select m in
100
These results give ~(.,%) = Y(.,%)/Y(o,%) X(',%) = Z(.,%)/Z(o,%) (= ~(.,%) of section 3) (= X(.,%) of section 3). (4.4)
As in section 3 m(.) and n(.) are meromorphic on C with simple poles only at the points {(n + ~)2 : n e N }. In particular these o functions are regular at O and we use this fact to construct the resolvent function $ as in section 3 above; in fact we can identify $ with ~ of (3.11a)
~(x,~;f) = ~(x,~;f)
but now defined for x (,l), % ~ C\{(n + i 2 :n e N o} and all 2) f ~ H2(l,l). It is convenient to define ~ : (;,I) H2(l,) > C by ~(x;f) = $(x,o;f); it follows that, see (3.11b), M[~(x;f)] = f(x) (x c (1,1)),
(4.5)
(4.6)
(4.7)
Now define a linear operator A on H2(l,l) by (Af)(x) = ~(x;f) for all f H2(l,l). (x (I,1)) (4.8)
We shall show A is a bounded, symmetric operator also that A has an inverse A I.
For this purpose we require Len~na (i) (ii) Proof (i) ~(.;f) ~ C[],] lirap ~, (';f)g = O +l (f H 2) (f,g E H 2) (4.5) and (4.6) of ~ and the
asymptotic properties of the solutions Y and Z of Legendre's equation. (ii) We note that if g E H 2 then Ig(x) l = Ig() + i.e. I< Ig()l + [Jo op]g,12 i/2 (x ~ I).
= o (l)
(x~
l).
101
Similarly at 1.
(f c H2).
(4.9)
(~;f)d ~ = Ix
X
M[~(~;f)] ~ (~;f)d
since f L2(I,I) (recall H2 c L 2) and ~ (.;f) ~ C[I.I], i.e. ~ (';f) c H2(f ~ H2). Also
if:
Hence
4f:
IIAfl[H < 411fIIN (f c H 2)
I n t e g r a t i o n by p a r t s g i v e s
(Af'g)H
{p~flg, + ~g} =
p~,g I
1
fl 1
M[~]g
= II M[~]g I = =
f' f'
f g (f.g ~ H 2)
1
f M[~(;g)] = (f,Ag)
Since A i s bounded i t
102
Suppose now Af = O, i.e. ~(x;f) = 0 then from (4.7) 0 = M[~(x;f)] i.e. f = 0 in H 2. = f(x) (x E (1,I)),
(x ~ (!,l));
Thus A ! exists.
Now define an operator S : D(S) H 2 H 2 by D(S) = {Af : f e H 2} and ! S f = A A standard to theorem H 2. f (f c D(S)) (4.10)
of ~, it follows
that if f e D(S) then f' e ACloc(l,] ) and so D(S) is strictly contained in H 2, even though the closure D(S) = H2; it may be seen that if S is bounded then D(S) = H 2 and this gives a contradiction. We conclude that S is an unbounded in H2(1,1); selfadjoint operator
spectrum with
O(S) = P~(S) = {(n + )2 : n No}; we note that this spectrum is identical with Po(T) of the operator T introduced in (3.6). of S, i.e. for some and = ~M[f].
(4.7), f = M[~(;f)]
section 2 this can happen only when the solutions Y and Z are linearly dependent, dependent {Pn(') i.e. when % c Po(S) on the corresponding and then the eigenvector f is linearly
: n e No}.
with corresponding
For all real numbers D @ Po(S) we can show that the range of SZI is H2(I,I), i.e. {(S  B I)f : f D(S)} = H2(l,l); this follows
of the resolvent
function ~ as defined in (4.5). set of S, and so O(S) = Po(S). operators in Hilbert form
see [I, chapter VII, now implies that the Legendre polynomials
103
a complete,
set in this space is then, say, {~n : n No} where ~n = (n + );/2P n (n E No). From this last result it may be shown that the Legendre polynomials are dense in those vectors of L2(I,;) which are also in and so we obtain,
H2(II,I); however this set is d~nse itself in L2(I,1) indirectly, yet another p r o o f ~ h e in L2(I, I). 5. Remarks on the operators T and S. completeness
operators T and S as defined in sections 3 and 4 respectively. T is an unbounded selfadjoint, differential operator in
L2(I,I) with a simple, discrete spectrum {(n + ~)2 : n E N o} and corresponding eigenvectors {Pn : n e No}. selfadjoint operator in H2(I,I) with we may hesitate to
The operator T, and its domain D(T), is defined directly in terms of the Legendre differential expression M; also we are able to give alternative and simplified descriptions of the elements of D(T), as
detailed in section 3. The situation for the operator S is different; we defined S as the inverse A I of a bounded, symmetric operator in H2(I,I). Whilst
we can say something about the elements of the domain D(S), in particular as sets of complexvalued functions on [1,1] we have D(S) c D(T)
(see the alternative definition T below), it does not seem possible to characterize the operator S directly in terms of the Legendre differential expression M. The definition of S in section 4, i.e. S = A I, depends
upon a general theorem in Hilbert space theory (see the result quoted in section 4 from [I]) which provides for the existence of S but, due to the generality of the theorem, general. cannot give a constructive definition in operator only in an indirect
sense in comparison with the operator T. It is of interest to note that we could have defined the operator T in the same way as S is defined in section 4. With the resolvent
function # defined by (3.]a) let the operator B be defined on L2(I,) by (Bf)(x) = (x,o;f) (x ~ (l,l))
104
to those used in section 4 we may prove that B is a symmetric, bounded operator on L2(I,I) into L2(I,;), and that the inverse B I exists; also that the operator T, as already defined in section 3, satisfies 1 Note that in the sense H2(I,I) c L2(I,I), we have A = B on H2(;,]). However, the inverse A I has to be determined in
T=B
H2(I,I) and B; in the space L2(I,I) so that there is no identification of S with T, even on D(S). It happens then that in the rightdefinite case, we are I = T, a
characterization which is not possible in the leftdefinite case. These remarks should also be read in the context of the general theory developed by Pleijel for both the right and left definite cases; see, in particular, the illuminating results given in the concluding remarks in [6, section 8]. 6. H alfranse Lesendre series. In the theory of Fourier series the two
collections of functions {sin n x : x ~ [o,~], n E N o } and {cos n x : x E [o,~], n e N o } give separate orthogonal sets in L2(o,~), both of which are complete in this space; these are termed halfrange Fourier series. The two collections conbined and extended to the interval
[~,~] give an orthogonal set which is complete in L2(~,~). The same phenomena is seen in Legendre series. The two
collections of polynomials {P2n+l (') : n e N o } and {P2n (.) : n e N o} give two separate, complete orthogonal sets in L2(O,I). The associated selfadjoint, differential operators are given respectively, say, by T o and T 1 where D(T o) = {f:[o,) C : f and fl E ACloc[O,l) f and M[f] E L2(O,I) f(o) = 0 and lim [f~] = O}
I
and
Tof = M[f]
(f e D(To))
(in the notation of section 3); and D(TI) and T; are given by replacing the boundary condition at the regular endpoint O by f4(o) = O. It may
105
of T
is given by
{(2n + )2 : n E No}; similarly for T! but with spectrum 3 2 {(2 n + ~) : n No}. There are similar halfrange Legendre expansions in the leftdefinite case.
106
References
I.
(McGrawHill, New York, 1953). 3. W.N. Everitt. Some remarks on a differential expression with an S~ectral theory and asymptotics of Mathematical Studies 13, 1974).
1328.
translated from the Russian). 5. A. Pleijel. On Legendre's polynomials New developments in differential Mathematical Studies 21, W. Eckha~s (ed), 1976).
equations 175180.
Annales Academiae Scientiarum Fennicae; Series A.I. Mathematica 2, 1976, 397408. 7. E.C. Titchmarsh. (Oxford) 8. On expensi~s in eigenfunctions II. Quart. J. Math.
ii, 1940, 129140. Eigenfunction expansions associated with secondorder I (Clarendon Press, Oxford, 1962). A course of modern analysis (University
E.C. Titchmarsh.
differential equations; 9.
I0.
On the mcoefficient of
Weyl for a differential equation with an indefinite weight function. Proc. London Math. Soc. (3) 29 (1974), 368384.
OF .AND
3DIMENSIONAL AT THE
POTENTIAL EDGES OF
Gaetano
Fichera
Icl raw, tory of Arthur Erd@lyi In which this have Let space terior smooth potential ditions
~z~ :x a = t on ,
paper been
we proved
shall [ iin ] in
extend the
to particular
more
surfaces a of cube.
the
resul
be the . We
the
domain the
X ~ of to A enough
conducting
surface function u of E
potential
function
is
continuous
in : R E
u E and
+
~y~ lira
+
~z ~(P) ~
= 0
in
: O.
Let
be
the
origin
of
X 3 and
the
unit
sphere
IP0
I = t . Let
5* tive
be
domain
of 5 we
; o
which shall
IPOI ~
does denote
not H R bythe
PO IPOI
coincide cone
with 5
If
[k
is
posi
number,
H R  IP If we =
R
the
~ S
IPOf
for
P~O
we
},
suppose that
denote
by 0 < R
~R
unit
bali
~ }%
B~ We
Ft R
for
~Ro
assume arcs of
that 5 ~ is 9 great
set of
formed them
circles,
point. be the We
~ R o
vertices of ~S"
of
, in measured ~h bythe
of 5 ~
o ~ I P  O I
by I h
ofX ~ defined
conditions
= tPOt(~hO
Let of
(1)
L~ consider
be
the the
LaplaceBeltrami eigenvalue
L~
f ),~ : 0 ~r ~ 0
operator problem
in on
S
on 5
In
the
domain 5"
we
95
108
The i) of tive
results of
can the
be
assumed
to
be
functions and 2)
tends +
multiplicity is is
equal
tothe
finite. in 5~
problem has
continuous the
correspondin by 5"
eigenfunction Let ~4 H R be
negative) H~ defined
the
the
conditions
P0
, IPOl ~ OS
for
P~o
H R pointing
Let towards E
be We
the
unit
normal in consider
~CQ) :  _L
a regular the
~___! ~
point the
of
shall
densityof
.
electric
charge in
The
following and
theorem
describes
the b e h a v i o u r ~ near
of
the
field edges
grad u
of the e l e c t r i c
density
the v e r t e x
of
~ 1 4 RI.
THEOREM
For
H~
(i~
u .....
~ i~)
(o < R
< R~
) we
have
(2)
where
a< :
grad
~:
r_,
IPOI
o~
IPOI
Ira ~0 k)
if
~h > ~
~:h(~)
I~
~h i log
if
k~h :
Im
 u~ h I
if
~h 4
(~)
The
symbol
is
the
Landau
capital 0
109
asymptotic G9 in it
cannot replaced in
be
improved if x
in tends
the to
sense an 0 of . ) ar
that
o by (z) a suitable
bitrarily If and
(3)
fixed Q ~ is ~ FI~
a point ( I h
0 m h ~I,+~ i ~ ) we have
u [h+,)
= 0
( I~
[~:
3(0)
PROOF. For 0 <
~:~(~;~:,,~,~'~)
'~,~ ~ ~ : , ) '
R o
a) 5" ,
we
have
The of
convergence We
~2U
of have
'~'U :+
the
series
must
be
understood
in
the
sense
L'
(5')
~U
~ 
Multiplying
both
sides
~rby and k
integrating
over5 ~
we
have
(4)
where
Zrk L,(oUdo)
(5)
The in are apply the differentiations right hand because Green side
get
permitted the
Hence
a_ _ = U.
~ __ d U Z
_ __ ~
U.
=o
and
therefore
UK( ~):
whe re c
K
c k?
+ %(:
and
and
c k
are
constan~
(~)
The
symbol
is
the
Landau
small o
110
Since U(q~)
is continuous as ) 0 ~
, we
must
have
and (6)
formal]y grad
U(qco)
k:~
grad~
zrk(co )
{, I g r a % ~
(7)
I grad
Mapping
?
the
u~(~)l
: I% ~
20(
~
a
+ ~
bounded
lemma
~ g rIa d
planar
II ]
~ I
domain
and
5~ o n t o
applying
d o m a i n by a s t e r e o (~)
of , [Iwe have
graphic
projection
(8)
l~k(~)l
~_ c 1~ ,
q
(9)
where
grad
C Set is
"G I z K
a
sup
constant
~h(~ )
: l~cohl
2
log
I~  coal if
Ph ( ~
"
From
(8), (9)
we
deduce q
(io)
where ~ Let is ~
Ivk(~)l ~ F ~ k ~ ~h(~)
a constant be
q
only
depending such
q
on for
~ co ~ 5 ~
a constant
that
h= ~)2 then
(12)
.............. ~
(f
tU(R
~)i~d~
_z R~ ~
(3)
See,in
particular,
(1.7),(1.8),(1.9),(i.
I0),(I.II)
of [I].
111
From
(ii), (12)
we
deduce
z
It p.442) Therefore,
any k
C~
asymptotic It follows theory eigenvalues of  that lim ~ k constants and po
I 3
z
(see ~ that 0 . for [2]
is that
known
from lim ~ ~k ~ k
 q~
there
exist
positive
p~ such
P ~.
We
have
for 0 4 ~
~R
Jim
ki k s {~q )PJ~ :
lira
exp
~~ ( {
logk
+ Po l o g
) = O.
J~
which proves the total convergence of the series
Ro
in is
the
interval
[0, R ].
Hence
the
development
of
grad~
given
by (6)
~ ) , assuming E
k:~
~ ~ ~
4
,
 
~z~ (d
~'+C
If
the
constant
is
such
that
for 0 ~ ~ ~
~R ~
ut I

<
6 ~ 11 ~
h:~ h
(~)
'
~(~)]
0P~: and
(forP'P~)
. In U<0 in fact in the ~ ~ . On
I~ ~01 <
development other
~
(6)
the
hand
estimate Hence
(13)
exists
that
112
a~ q
(lap
for
_P~
l[grad~I~l
,_ 8~
,
]7 r (co)
h:
{
HR(I~u...~I~)
We may s u p p o s e
IP@l
<
8~,
_P we h a v e
o < c~ _~ ~..
For such
oo
~
lgrad
ICi~l I g r a d ~ ( a ) )
,2
[c,l la~
e( ~
%(c~)I
2.
20(  .2
+ ~
~Y (co
_ o__~?,=h
Let M~
(~),,r:z :x,,(~g
be a positive
* ~ )
such
< ~.
that
"
constant
for ~ ~ 0
(~5)
~
Hence
~
for
~ ~"
)'
(13)',(14),(15)
I,.:~ h
]grad
oa~r~(a))[ M t 6 I
(~
h.:~
g "~i (co)
~(co)
V1 ~(2a)
11 =~,(~)].
h:l
The re f ore
grad
Since lemma III If we have Q P~ ~ of = ~w
I.tr4(co)l z Ic,I
0 and ~(a))>O in ~, the S
.
last estimate contradicts
~ (~)...
.. V9 (~)
" L obtain
where the
is
Ic~(G)I I[ grad"]QIwe ~
estimate THEOREM
and be of
condition
for
a~
C*(HR}
( 0
< P, 4 PROOF.
R o)
(h={.
that
greater ,
smallest be easily
for the
hemisphere
equals From 2.
inequality
~i > 2
113
we is deduce ~ > ~ .Hence in FI R necessity and from the proof on the of Hg ~
i
Theorem we I
h ( h ~
see
~ , " "' ' @
that
)
grad
continuous The
vanishes from
follows
fact
that
the
be
we is
that
for
at
least
one
value we of h
have
Mh > ~ THEOREM
smallest
J , . j
of
2 IX~
the
numbers
2 1"1
that
are
positive.
Then I g r a d u l
e LP(H~)
( 0 ~ 1~ ~ Ro ) f o r
any
such that
4 a p ~ ~. We h a v e
finite:
PROOF.
integral is
f r o m (2)
that
} g r a d ~ l c: L P(Ft R) i f
the
following
FIR
On following the other hand are this integrals
h~
integral is finite if and only if the finite:
where ~h > ~T
the
integral remark p be
be proof
for h
any such
that
THEOREM
Let
smallest
numbers
that ( 0 <
are R PROOF.
Then p by
density ~ p .
~(Q)
belongs
to
LP(g4H
sector ~(Q)E
which H LP(~ if R)
is
the the
intersection follow
of ing
H R
with
integrals
finite:
(16)
for any h such
( c u ) J P d Y'~
one of the following inequalities holds:
h > ~ '~h#~ > ~ " Assume in the pole only 0 if and for polar all
plane 0 ~0 h ~oh+ ~
)
.The that
114
this IV
the
proof a
improves
from ) paper
theory
developed
states
that ~
the
of
the is
rigorous ) :
results
been
for ([4],
0.62153
0.68025
which l e a d t o t h e f o l l o w i n g bounds f o r
0.4335 These ing bounds together with < ~
~
<
:
0.4646. III and IV lead to the follow
theorems
function of a
theoretic cube
results
6000
for
the
electric
field
and
the
electri
density
where
is
an
arbitrary
number
such
D ~ a < I. that
[ 1 ]
G. FICHERA, Asymptotic behaviour of the electric field and densit of the electric charge in the neighborhood of singular a points conducting surface, Uspekhi Mat. Nauk,30:3,1975,pp. lO5124; English translation: Russian Math. Surveys,30:3,1975,pp.107127; Ital.translation:Rend, del Seminario mat.dell'Univ, e del Politec. di Torino, 32,197374,pp. II~143. Methods York, 1953. of Mathematical Physics,vol. I; Inter
[ 2
] R. COURANTD. HILBERT, science Publ. New ] M.A.SNEIDER, s_aa, Mem. Acc. G. FICHERAM.A. le voisinage Paris,278
[ 3
Sulla capacit& elettrostatica Naz. Lincei,X,3,1970,pp.99215. SNEIDER, Distribution des sommets et des 1974,pp.13031306.
di
una
superficie
chiu
[ 4 ]
de ar~tes
~lectrique C.R.Acad.
dan Sc
A,
SINGULAR PERTURBATIONS
OF ELLIPTIC
P. Habets
. Introduction A major contribution boundary value problems [5]. An extension A. van Harten bibliography). to singular perturbations of linear elliptic
(BVP) is due to W. Eckhaus and E.M. de Jager equations has been worked out by for a complete
to nonlinear
[11][12]
As a first step, we consider the nonlinear domain ~ C ~2 eL2u + Liu = f(x,u) u = g(x) where E > O, u E ~, x E ~2, Li is a linear operator; in
elliptic BVP in a
[I],
existence of solutions of (1.1) and a convergence solution of reduced problem L1u = f(x,u)
u = g(x)
in ~
(i .2)
on F
is
of order E except
in an arbitrary
small
boundary
small neighbourhood of a set D where the solution (1.2) is not C 2. We do not suppose the set
to be convex, which allows free boundary layers. Also, free boundary layers appear along the parts of the boundary of ~ which contain characteristics of the reduced problem (1.2). However, in such cases,
we give no informations
on the behaviour of the solutions of (I.I) in layers. Our work is very much
a neighbourhood of these free boundary in the spirit of A. van Harten (I.I) is supposed
solution of (1.2). This allows the use of the maximum principle together with a constant as a barrier function. In our work, we
116
using monotone
iteration.
In a second part, we allow nonlinearities More precisely, we consider the BVP EL2u = f(t,U,Ux,EUy) u = g Using a theorem of M. Nagumo
in on ~ inequalities we
results of the first part. More general [2]. The main drawback of this
using H. Amann
u(o) = u(]) = n"(o) = u"() = 0 which can be interpreted to high order equations worked out in [6] L.F. Shampine [4]. as describing a beam with pinends. This extends [3] and
2. A fixed point theorem for increasing maps 2.1. Let ~ be a bounded domain in ~ n and C(~)"the Banach space of continuous
functions u : ~ ~ ~ together with the norm Iluil = sup lu(x) l. The space C(~), IE.[i together with the order u] ~ u 2 iff Vx ~ []] ul(x) ~ u2(x)
[~,8] = {x ~ C(~)
~<
x < $~.
if u < u 2 implies
T(u l ) ~ T(u 2)
1t7
2.2. PROPOSITION
and
u_~ = T ~ _ 1 Uk = TUkI
3.
3. I. Let ~ C fl~2 be a bounded domain whose boundary C 2+v manifold for some v @ (o,I). Consider
the differential
L 2 =  all(X,y)DxDx  2a 12(X,y)DxDy  a22(x'Y) DyDy L 1 =  p(x,y)D x + q(x,y) where the following assumptions (Hi) aij , p, q E cV(~); are satisfied :
(Hiii) V(x,y) ~,
where E > 0. It is well known [l] that under assumptions (Hiii), one can define the solution operators K : cV(~) ~ c2+V(~), R : ~+v~)~ c2+V(~), f ~ Kf g ~ Rg
where Kf denotes the solution of the BVP (3.1) with g m 0 and Rg the solution of (3.1) with f m O. The solution of (3.1) reads then u = Kf + Rg.
118
~ ~ be such that
fu E C(Hx~), g E c2+V($~);
V ( x , y ) e ~, u t > u 2 implies
3.3. LEMMA
3.4. We are now ready to prove the main tool of this section. THEOREM
Bj E C2(~), j = 1,2,.,,1 such that for any i and j, ai ~ Bj, L2a i + LI~ i ~ f(.,a i) , L26 j + LIB j ~ f(.,~j)
in H,
(3.3)
such that
119
Proo~
Let u = Kf(.,~ i) + Rg. Then (sL 2 + LI)(~ i  u) = eL2~ i + LI~ i  f(.,~i ) < O, in ~, ~.  u = ~.  g ~ O, on ~ ,
I i
principle [lO]
a.  T~. = ~.  u ~ O
i i l
Further
This proves
T~
~ = max a i. Similarly
From Lemma 3.3 and Proposition we only know uk ~ u in C(~). from Schauder's estimate I~Ic2+~ the sequence there exists converging C2
uk ~ u.
follows, in C2(~),
except notice
that that
To prove
subsequence
[u.,] converging in C2(~). At last any 2 K ukC~ v converges in C(~) i.e. v = u. This implies
3.5. Consider
the reduced
I 3 h > O
: [x  h, x [ {y} C ~ }. shows right away that the solution at points on horizontal lines
of simple
examples
singularities
120
Choose
~ > 0 as small as we wish and define ~o @ C2(~) u(x,y) = u(x,y) if (x,y) E ~ \B(D,~).
such that
3.6. Consider
next a set
= l(x,y) : a < y < b , ~ (y) < x < ~(y) }
C ~ \
B(D,6)
(3.s)
n ~
= {(x,y)
I a < y ~ b,
x = ~+(y)
and
e C2([a,b]).
&
&
fig. 2. example of a set ~o C
Define
O E C~(~) = ] if if
p(y)
p(y) = 0 p(y) @
= ~_(y)
Consider
constants
121
Let us choose A large enough so that y E [a,b], x # ~+(y) implies ~(x,y) < uO(x,y) and C such that y ~ [a,b] , (x,y) @  A ~ g(x,y) ~ implies
(x,y) E ~ ,
Let I be an interval large enough such that for any (x,y) E ~0 (x,y) A* l C*(d  c) @ l with A* = sup {I:  g I : (x,y) E ~ , y E [a,b]} , ~ , y ~ [a,b]} and cL2~ + L1a < f(.,a) we shall
(d  x) , (x,y) E
}. In order to prove
(Hvi) for some ~ > 0 and all (x,y) E ~ p(x,y) k ~ , One computes
, u E
q(x,y)  fu(X,y,u) k 0
~xx = USx
~xy = U~y + %~Y' AeXl(x  Y) + CO' ~yy = U~y  (%lY" + %~X'2)Ae %l(x  X)  CO"(H  x)
e2a < e2u + K[(%~ + hl)Ae %l(x  Y) + %~gBe %2(d  x) + C] with K larger than : sup (all  2a12 Y' + a22 ~'2) , sup (a22X") , sup all ,
sup (2a120' + a220"(d  x)). Next, gL2~ one has + LI~  f(x,y,~) < gL2: + K[(%~ + %l)Ae %l(x  Y) + ~2 ~ %2(d  x) + C] g~e~2 _ P [:~ + %lAeXl(X  y) + %2~BeX2(d  x) + CO]  f(x,y,:0) + q:O  X) + CB(e%2(d  x) _ I) + Co(d  x)]
_ [q _ fu(X,y,e)][AeXl(X
122
with 0 E [a(x,y)
, uO(x,y)]. 
Using assumptions
~L2~ + LI~
c]
~ [%iAe%l(x  y) + %2EBe%2(d  x) + CO ]
+ gB(e%2(d  x) ~ I).
Let us choose %1 and %2 such that P  ~K)% I = 0 EK%2  P%2 + ! = 0 i.e. %1 = ~ 1 and ! %2 = ]~ + O(). It follows f(x,y,u O)
gL2e + LI(~  f(x,y,~) ~< gL2u + L1u  f(x,y,u O)  (~  gK)C < O for g <~ ~/2K if (x,y) e ~o ~L2~ + L1a  f(x,y,~) ~< sL2u + gKC  CB < O for B large enough, and if y @ [a,h] , x ~< ~_(y) and C large enough ,
sL2(~ + LI~  f(x,y,~) _< gL2u + L1u  f(x,y,u O) + EKC  ~ 2 A < O for A large enough.
In a similar way, one proves that for appropriate constants, the function
choice of the
123
3.7. Theorem 3.4 and the construction 3.6 prove the following theorem, THEOREM
Let ~ > o
Assume assumptions (Hi) to (Hvi) are satisfied. Then the BVP (3.2) has a solution u such that for ~ small enough and any i
u(x,y) = u(x,y) + O(e  % I ( x  @  i (y)) + e) in ~. 1
where
o <
%1 =
0()
defined by(3.7~.
4.
Nonlinearities
with
Let ~ C ~2 be a bounded domain such that for each p E ~ , there exists a cone K with vertex p and a neighbourhood 0 of p such that (K ~ O) lies outside of ~. Consider next the differential operator
(Ai) for some C and all (x,y),(x',y') e ~ , i,j = 1,2, l aij(x,Y)  aij(x',y')l ~ C H(x,y)  (x',y')ll;
124
let f : ~ x ~3 ~
, u [p;+p]
4.2. Consider
result
is then a consequence
M. Nagumo [9].
THEOREM
' i = 1,2,,..
k and ~j C2(~)
En2a i ~ f(.,~i,aix,g~iy) a. ~ g l
Then if assumptions (Ai) to (Aiv) are satisfied there exists a solution u of (4.1) such that
max a . ~ u < min BJ i i j
problem
h > o
: Ix  h,x[{y}
\ D). We choose
~o C2(~)
such that u(x,y) = u(x,y) if (x,y) ~ \ B(D,~) of 3.6 we consider a set ~o and
Using a function a
the notations
and assumptions
= ~o _ Ae~1(x
 y(y))
_ ~B(e%2(d
 x)  I)  Cp(y)(d
 x)
125
Let I be an interval defined as in 3.6 and let us replace assumption Hvi by (Av) for some ~ > o , Q > o a n d all (x,y) ~ , u I , (v,w) ~ ~2
 Q ~ fu < 0 , fv ~ ~ ' ~fwI ~ Q ,
+ gK[(%~ + %1)Ae kl(x  Y) + A2g~e2%2(d  x) + C] ~  f(x,y,~O,~Ox,O ) + fu(X,Y,Sl,e2,@3)[Ae%l(X _ fv(X,Y,el,e2,83)[%iAe%l(X  y) + cB(e%2(d  x)  I) + Cp(d  x)]  y) + %2sBe%2(d  x) + CO]
+ Sfw(X,Y,el,62,e3)[%iY'Ae %l(x  Y) + Cp'(d  x)  ~Oy] with (el,e2,e3) a point on the line segment joining the points (Av) and if %1 > O and (~o, ~ o , o )
%2 > O are
gK% 1  (~  gK  ~Q sup :y'i) = 0 EK%~  ~%2 + l = 0 one has gL2a  f(x,y,a,ax,Cay ) < EL2~  f(x,y,~O,~Ox,O)
 ~%IAe %(x  Y)  EB  pO___CC2
At last, using the type of argument used in 3.6, one proves EL2~  f(x,y,~,~x,E~y) <0 0
Similarly,
it can be
shown that the function 6 = ~O + Ae%l(X  y(y)) + EB(e%2(d  x) _ I) + Cp(y)(d  x) is such that ~L26  f(x,y,6,6x,e6y) 6 ~ g ~ 0 on 3~ in ~ ,
126
4.3 prove
the following
theorem,
with ~i and ~+i E C 2. Assume assumptions (Ai) to (Av) are satisfied Then the BVP (4.2) has a solution u such that for g small enough and any i
u(x,y) = uO(x,y) + O(e%l ( x ' ~  i (y)) + ~) in ~i '
where
o < ~1 = O ( I / E ) .
5.
The beam string problem of Proposition 2.2 is that it can be applied consider to higher
As an example
= f(t,u)
= o = o (5.1)
= u"(1) = u (2)
p :[0,I] ~
(i) p is continuous and for some p > o and all t @ [0,I] , p(t) ~ V 2 ; [0, I] x ~ ,
(ii) f and fu are continuous and for all (t,u) e fu(t,u) > O.
5.2. To solve
the reduced
problem
 p(t)u"
u(0)
= f(t,u)
(5.2) = o,
= u(1)
let us define
where Ko denotes
the solution
127
 p(t)u" = f(t)
u(O) = u(1) = o .
From the maximum principle the Nemitskii operator V : u ~ Fu = f(.,u) . (5.3) [I0] , K c is increasing. Next we introduce
It is then obvious that solutions of the BVP (5.2) are the fixed points of the increasing operator
To = KoF.
Their existence can be obtained as follows
5.3. PROPOSITION
, So(1) < o
and
~k = To~kI
= roukI
converge to solutions u and u of the reduced problem (5.2) such that % Proof <$ < _ u < u < Uk< SO .
theorem TO is compact on [c~,S]. The inequalities
Prom ArzelaAscoli's
.(Toao.  ~o ~< O , To~o(O)  ~o(O) i> O , To~o(1)  as(1) > 0 and the ~''
maximum principle imply To~ O > ct O. Similarly, one proves ToS o ~< SO U and Proposition 2.2 applies.
5.4. To study the BVP (5.1), let us introduce Kg : C([O,I]) ~ C([O,I]) where KEfdenotes
the operator
, f~K~f
128
By applying twice the maximum principle, it is easy to show that f(t) i> O implies u" < O and next u ~> 0, i.e. the operator K E is
increasing. Hence solutions of the BVP (5.1) are the fixed points of the increasing operator Tg = K~F. Once again, ArzelaAscoli's theorem imply T E is completely continuous and
5.5. PROPOSITION
E C4([O,I]) such that g2siv  p(t)S" < f(t,s) , E2B iv  p(t)B" ~> f(t.$) , O < t < 1 ,
W'(o) s (0) i> o < 0 , W'(1) , s (1) i> o ~< 0 , B"(o) , B (0) ~< o ~> 0 , S"(1) , IB ( 1 ) ~< o > 0 , .
and
TB ~< B
5.6. THEOREM
such that
Iv

So(O)
<o
%(1) ~<o
Bo(O) />
Bo(l) >/
o.
Then, for ~ small enough there exists at least one solution u~ of the BVP (5.1) such that
a o ~< u + O(E 2) ~< flo "
129
Proof.
ut ~o + c2 ( A e  ~ +
~(lt)
Be" E  C)
where A,B and C are positive constants chosen such that ~"(O) = ~o(O) + ~2A + ~2Be '~/~ > O , ~"(I) = ~o(I) + ~2Ae~/c + ~ 2 B >~ 0 ,
(O) = ~o(O) + E2(A + Be ~/~  C) ~< 0 , (I) = ~o(I) + g2(Ae~/E + B  C) ~< O. One computes next g2 iv _ p(t)~"

iv f(t,~) = E 2~O + D 4 A e  ~
+ ~4Be ~(]~t)
,, Dt p 2Be_D(~t )  p(t)~ O  p~2Ae~ _  f(t,~o)  S2fu(t,~ O + ~(~  ~O))[Ae with ~ E ]0,I[. Hence, for some K > 0 and c small enough E2~ iv  p(t)~"  f(t,~) ~  p(t)~ O  f(t,~ o) + KE 2 < O. ~he theorem follows from Proposition 5.5 and a similar choice for the function B ~t _~(lt) B = ~O  g2(Ae~ + Be g  C).
a
~(1t) C] + B e  c 
References [I] H. Amann, Fixed point equations and nonlinear eigenvalue problems in ordered Banach spaces, SIAM Review 18 (1976), 620709. [2] H. Amann, Existence and multiplicity theorems for semilinear elliptic
boundary value problems, Math. Z. 150 (1976), 281295. [3] N.I. Bri, On boundary value problems for the equation cy" = f(x,y,y') for small e, Dokl. Akad. Nauk SSSR 95 (1954), 429432. [4] F.W. Dorr, S.V. Parter, L.F. Shampine, Applications principle to singular perturbation problems, 4388. [5] W. Eckhaus, E.M. de Jager, Asymptotic solutions of singular perturbation problems for linear differential Mech. Anal. 23 (1966), 2686. equations of elliptic type, Arch. Rat. of the maximum
130 [6] P. Habets, M. Laloy, Perturbations singuli&res de problgmes aux limites : les sur et soussolutions, S~minaire de Math~matique Appliqu~e et Mgcanique 76 (1974). [7] F.A. Howes, Singular perturbations and differential inequalities, Memoirs A.M.S. 168 (1976). [8] M.A. Krasnosel'ski, Positive solutions of operator equations, Noordhoff, Groningen 1964. [9] M. Nagumo, On principally linear elliptic differential equations of the second order, Osaka Math. J. 6 (;954), 207229. [10] M.H. Protter, H.F. Weinberger, Maximum principles in differential equations, Prentice Hall, Englewood Cliffs, N.J 1967. []I] A. van Harten, Singularly perturbed nonlinear 2nd order elliptic boundary value problems, PhD Thesis, Utrecht 1975~ []2] A. van Harten, On an elliptic singular perturbation problem, Ordinary and Partial Differential Equations, Ed. W.N. Everitt and B.D. Sleeman, Lecture Notes in Mathematics 564, pp~485495, Springer Verlag, Berlin Heidelberg NewYork 1976.
by
F. A. Howes* School of Mathematics University of Minnesota Minneapolis, ~innesota 55455
and R. E. O'Malley, Jr.* Department of Mathematics and Program in Applied Mathematics University of Arizona Tucson, Arizona 85721
I.
Problems with boundary layers at one endpoint Many physical problems can be studied as singularly perturbed twopoint vector
OJt_<
(1)
y(0), y(1) prescribed
where
is a small positive parameter (cf., e.g., Amundson (1974), Sethna and Scalar problems of this form are analyzed An enlightening case
history of such analyses was given by Erd~lyi (1975), and important early work includes that of Coddington and Levinson (1952) and Wasow (1956). For simplicity, let us assume that in y and t f and g are infinitely differentiable e as
E O.
We'll first consider the vector problem under the assumption that the
reduced problem
(3)
f(uR(t),t,0) > 0
there (i.e.,
f
Supported in part by the National Science Foundation under Grant Number MCS 7605979 and by the Office of Naval Research under Contract Number N0001476C0326.
132
uR
expect to have
t = 0, ~ 0.
or for a constant
(4)
I ~ f(uR(0)+s, 0
0, 0)ds > 0
for
~ + UR(0)
UR(0)
and
y(O) llzli =
il~li~
We note that if
is the gradient
VF(z  UR(0)) ,
equivalent
to the condition
is then pathindependent.
Indeed,
the
used by Howes for the scalar problem and it is weaker than that f(y,0,0) > 0 for all y.
y(1) = UR(1)
Figure 1
The results of Howes and others have a solution y(t,c) of the form
(i) will
(5)
expansion
(6)
U.(t)s j
j=0
133
t > 0,
0o
(7)
~I(T,E)
Z K. (T)c j j=0 J
(8)
= t/E
tends to infinity.
Under weaker
higher order terms or complete boundary layer behavior, but they can easily be generated. Applying his results to the boundary value problem for the remainder
terms, however, shows the asymptotic validity of the expansions so obtained. The outer expansion (6) must provide the asymptotic solution to (i) for t > 0, since ~ is then asymptotically negligible. Thus, the terms U. can be 3 successively obtained by equating coefficients in the terminal value problem
(9)
U(I,B) = y(1).
Evaluating at
c = 0,
U0
(i0)
f(U0(t),t,0)U&(t) + g(U0(t),t,0) = 0,
U0(1) = y(1)
UR(t)
Succeeding terms
Uj,
(ll)
f(U0(t),t,0)U](t) + fy(U0(t),t,0)Uj(t)U&(t)
+ gy(Uo(t),t,O)Uj(t)
= hj_l(t) ,
U.(1)] = 0
where
hi_ I
is known in terms of
t, U0(t),
..., Uj_l(t).
tion (3) implies that (ii) is a nonsingular initial value problem, so it also has a unique solution throughout ing the outer expansion 0 < t < i. with H Thus, there is no difficulty in generat
U(t,e)
(12)
134
 f(U(ET,e),ez,e)) ~
dU (e~, c )
+ g(U(~'~,e) + n ( z , ~ ) ,
~'v, e)
(12)
 g(U(eT,e),er,e) ],
T > 0
(13)
d2~ 0
dr 2
while later terms must satisfy linear problems d2~j (14) dT2 + f(U 0(0) + 17 0(T) dN. J 0, 0) dr ~j(0) : Uj (0) and their derivatives T and HO(T). The
d~ 0 + fy(Uo(O) + nO(Z), O, O)Hj(z) ~ = kj_l(T), where dN~/dT, kj_ I is a linear combination of preceding terms with coefficients that are functions of ~
% < J,
d~ 0 dY +
fT
(15)
dn 0
f~0 (T)
T ~0'
f(U0(0) + w, 0, O)dw,
~ i 0,
d~ : J0
Multiplying by the boundary layer stability condition (4) implies that
~0(z)
(16) 1 d 2 dT ]l~0(T)l]2 = "H~(T) I 0 H0(T) satisfying f(U0(0) + z, 0, 0)dz < 0 [I~0(T)}] ! fly(0)  UR(0)H = IIH0(0)[I. ~0(r)~ will decrease monotonicof (15) at T = ~. t = 0) implies that the H0(T) = 0
Thus, our boundary layer stability implies that Ultimately, than some ~0(T) K > 0
will become so small that (3) (for and (15) then implies that
eigenvalues of
(17)
H0 (T) = 0(e<T),
135
i.e., Y0 is exponentially decaying as T + =. Although we can seldom explicitly
integrate the nonlinear system (15), we can approximate its solution arbitrarily closely by using a successive approximations procedure on (15) (1964)). wise. Knowing ~0' we next integrate (14) for j = 1 (cf. Erd~lyi
Rearranging
j(T) =
fco
and
dK 0 d~T . (r)H.(r)]3 + kj_l(r)}dr is known whenever equation ~. 3 dH0/dT commute. Thus, ~. 3 satisfies the integral
IT
(18)
where P(~) H.(T) = 3 P(T)U.(0) 3 P(T)Pl(r)%j(r)dr 0 is the exponentially decaying fundamental matrix for the linear system
T > 0,
H(0) = I.
In general,
though it
directly provides the solution of (14) when the commutator We note that the boundary layer jump the minimum value of II~II > 0
(19)
The jump can, in practice, be quite large. ple, if f(z,0,0) > 0 for all z
It gives precise limits to the jumps for certain scalar problems reconsiders an example of 0'Malley (1974)).
(Howes (1977)
(20)
f(uL,t,0)u L + g(uL,t,0) = 0,
136
(21)
f(uL(t),t,0) < 0
for
0 < t < 1
(22)
eT
fe
0
for all
8 + UL(1)
on paths between
UL(1)
and
y(1)
satisfying
0 < IlOlf
IIy(1)  UL(1)il. Nonuniform convergence of the solution to (i) would then take place near t = i, depending on the stretched variable
o = (i
t)/E,
0 j t < 1
would be
uL(t).
If
were nonsingular
with eigenvalues having both positive and negative real parts along an appropriate solution of the reduced system, we must expect boundary layer behavior near each endpoint (cf. Harris (1973) and Ferguson (1975) for discussions of problems where f (y,t,0) E 0). Y 2. Problems with boundary layers at both endpoints Let us now consider the "twin" boundary layer problem
(
(23) ~ ~Y" + g(y,t,e) = 0, y(0), y(1) prescribed 0 < t < 1
of
(24)
g(u,t,0) = 0
U0(t )
throughout
0 < t < 1
(25)
gy(U0(t),t,0) < 0
there, i.e.,
gy
(U0(t),t,0) ,
0 < t < i.
Motivation for this assumption is obvious if one considers the linear scalar problems with cy" y = 0, while generalized stability assumptions are sometimes
appropriate and necessary (cf., e.g., Howes (1978) or consider the scalar problem with g = y2q+l). With (25), one can hope that a solution to (23) exists which within (0,i). Since we won't generally have either
converges to
U0(t )
137
U0(0) = y(0)
or
U0(1) = y(1),
we must expect
layers t = 0
of thickness
(cf. Fife
(1975),
(1978)) suggests
limit the size of the boundary layer They'll certainly be guaranteed (cf. Kelley (1978)). if
[[y(0)  U0(0)I;
ily(1)  U0(1)[[.
Indeed, cient.
(25) is suffi
(26)
~
where
U,
9,
and
and the terms of the left boundary layer correction stretched variable
(27)
p = t/~TEe
Q + 0
as
(28)
o = (i  t)//~e
(29)
U(t,E)
Z U. (t)g j j=0 j
should therefore
satisfy
(30)
eU" + g(U,t,s)
= 0,
0 < t < i
and converge
to the solution
U0(t)
(31)
where
coeffiof (24)
138
would,
of course,
result
in different
sequences
of perturbation
terms
U., J
j > 0,
stability
function
holomorphic
the outer
If we set
(32)
we convert the problem
y(t,~)
= U(t,e)
+ z(t,g),
problem
(33)
cz" = h(z,t,~),
0 < t < i,
z(0,~)
= y(0)  U(0,s).
Here
~
h(z,t,e)
= ~U"(t,s)
 g(U(t,)
+ z,
t, S)
satisfies
(34)
h(0,t,c)
= 0(~ N)
N > 0
since
EU" + g(U,t,e)
= 0(eN).
In particular,
the reduced
system
h(z,t,0)
= 0
corresponding trivial
problem expansion
(33) has
unique)
solution
Henceforth,
(33) and,
corresponding
of the form
(35)
z(t,c) = v ( p , ~ )
+ w(u,~)
providing
boundary
layer decay
to zero within
0 < t < i.
Our smooth
ness assumptions
be required
in a domain
~EI,6
= {(z,t,e):
0 < t < i,
0 < ~ < I} ~
where
sI
is a small positive
number and,
for any
8 > 0,
we define
i d6(t) =
llz(0)  U0(0)II 6,
+ 6,
0 <
t <
~ < t < 1  8
139
l
of [[zll = ~ z T z . Here (36) 0 < t < i, where
ilz(1)  U0(1)li
+ 6,
i
~ < t < i.
llz(0,a)II and
dt 2 = 2(z")Tz + 211z'il 2
llzll. Further,
(37)
llz'[i > 2
(38)
0 < t < i.
(Through the inequality (37), then, we eliminate the first derivative term from (38). We note that (38) is an equality for scalar problems.) We'll now ask that for all function (z,t,s) in ~ gl,~' there exists a smooth scalar
~(n,t,g)
such that
(39)
where
140
(40)
(O,t,c) ~ O,
(O,t,O) = O,
(O,t,O) > 0
(41)
and
I
cifically, endpoints.
~ ~(s,l,0)ds > 0 0
whenever
0 < ~ ~ Itz(l,O)li
if
z(l,O) O.
Spe
implies the stability of the trivial solution of the while (41) implies boundary layer stability at both
(42)
where
m(t,c)
(43)
cm" = (m,t,e),
0 < t < I,
m(O,c) = iiz(O,c)ll,
m(1,g) = i~z(l,c)II.
The bounds (42) follow from the elementary theory of differential inequalities since zero is a lower solution for lizil and m is an upper solution (cf. Nagumo Further, ~(0,t,0)
(1937), Dorr, Parter, and Shampine (1973), and Howes (1976)). = 0 and ~ n (0,t,0) > 0
~(m,t,0) = 0
is the appropriate hypothesis for the needed boundary layer stability of this solution. Indeed, the solution of (43) satisfies
(44)
where
r0
d2r 0
(45)
while sO
dp2
= (r0,0,0)
P > 0,
141
The solutions to (45) and (46) are easily shown to exist and be unique. ing (45) by that dr0/dP, for example, and integrating from p
Multiply
to infinity implies
(dr0)ir0
d7=
~(s,0,0)ds
>
(by (41)).
Thus,
r0
(47)
dr
dp r0(P) r0 = 0 at
S (p)
0
~(s,O,O)ds,
Hence,
rest point
(0,0,0)s
~_i (0,0,0) > 0 implies that the decay of r 0 to zero is exponential as p ~ ~ Sn (When r0(0) = 0, we have r0(P) E 0 since there is no need for a boundary layer correction.) Continuing by solving linear problems, we could obtain an
(48)
m(t,s) = r ( p , ~ )
+ s(o,~).
In terms of the original problem (23), our stability hypothesis (39) becomes the inequality
where
expansion (26) for an asymptotic solution for the vector problem (23). Now, we return to the vector boundary value problem (33) and its asymptotic solution in the form (35). totically negligible (o Near t = 0, w and its derivatives should be asymp~
being infinite), so (33) and (35) imply that the initial v should be a decaying solution of the nonlinear
(49)
pp
= h(v,~p,e),
p > O,
v(O,~) = z(O,e).
(50)
by substitution into (49).
v(p,/T)
142
problem d2v0 ......h(v0, 0,0) do2 v., J j > I, d2v. dP 2~ = hz(V0,0,0)v j + dj_l(p), (52) v.j(0) = 0, j odd; p _> 0 j even
(51)
p ! 0,
v0 0
as
O * ~.
Later terms
v. + 0 3 where dj_ 1
as
p ~
llv011p0 ~ ~(llv011,0,O),
@ ~ 0,
llv0(O)ll = r0(0),
v0(P)
such that
(53)
0 ~ llv0(P)ll j r0(0),
P ~ 0
(and
v0(P) ~ 0
if U0(0) = y(0)).
No explicit solution
v0
can be provided,
= hz(0,0,0) > 0 (whose eigenvalues have strictly positive real parts by our stability assumption (25)), variation of parameters can be used to express the solution of (52) in the form
(54)
2 0 J
(r)dr +
f
p
where
fashion, we could generate the terms of the terminal boundary layer correction w(o,~) of (35). Thus, we've formally obtained (35), which we expect is a locally
143
on
~ Thus,
automatically if we take
hold if
gy(y,t,0)
or
h (v,0,0)  I > 0
z
y > 0 that
and all
satisfying
0 ! livJi
hT(z,0,0)z
= zTh (~,0,0)z > yilzll2 g Thus, taking ~(n,0,0) = yn, both hold. ~ (0,0,0)
point > 0
~. for
0 < D ~ Jlz(0,0)ll
sx" = F(x,x',t,s)
small.
Thus, Kelley
(1978) considered
problems where
(1968) considered
3.
Examples a. A problem with an initial boundary Let us consider the vector equation layer
gy" + f(y,t,c)y'
+ g(y,t,~)
= 0,
0 < t < 1
where
y =
lyll f (yll
Y2 1 solution Y2 uR f(uR,t,0)u ~ + g(uR,t,0)
YI+ , and g = 
i ).
= 0,
UR(1) = y(1)
we must require
uR
to be stable in
0 < t < ],
i.e.,
f(uR(t),t,0)
< 0
144
t = 0,
f~
0
for all
such that
0 < lJCJl ! fly(0)  UR(0)ll. the reduced problem has the solution
More specifically,
UR(t) = ( D t C t +
where
= URl(0) = i + Yl(1)
and
D = UR2(0) = i + Y2(1).
Stability of
uR
C + D > 0
and
CD > i,
i.e.,
Yl(1)Y2(1) > Yl(1) + Y2(1) Further, boundary layer stability requires that
2.
s( wl+c w21l(dwll
0, 0 i.e., 2 + 2D~ 2 > 0 1 + D dw 2 ~I + 2C~ 3 + 4~i~ 2 + ~ for all ~ = 2 satisfying 0 < ll~ll = lJy(0)  UR(0)ll = y(0) (Yl(0)  C) 2 +
are thereby restricted to a circle about II~II of the nontrivial zeros of the ~ will satisfy
with radius less than the least norm Setting ~2 = t~l' such a
cubic polynomial.
(i + t3)~l = 2(C + 2t + Dt 2)
145
~i = O,
t = =,
is
2D.)
y(1) = (~),
corresponding
tmi n = 0.291.
limiting solution of our twopoint problem is provided by in the circle of radius 3.390 about (~).
estimate for the "domain of attraction" of the reduced solution that boundary layer stability need only hold for tory joining y(O) and uR(0). ~ + UR(0)
analogous to the simplest cases occurring in the analysis of solutions of the scalar problem ey" + yy'  y = 0 (cf. Cole (1968), Howes (1978), and elsewhere).
b.
A problem with twin boundary layers at the endpoints Consider the vector problem
gz" = h(z,t,~),
0 < t < i
hz(0,t,0) = ( II
ii )
i i i. ~
such that
hT(z,t,e)z _>
~(lJzll,t,s)IEz]l.
Here 2 2 4 4 hT(z,t,e)z = (zI + z2)  (z I + z2) _> IIzJl2(l  11z~2). 4 4 zI + z 2 < 2 2 2 (z I + z2) , so we can take ~(n,t,~) = n(l  n2).
Since
Clearly,
#(0, t,e) ! 0,
(0,t,O) = 0,
~n(0,t,0)
> 0
and
146
0 t
~(s,i,0)ds
= ~ n2(l  n2/2)
> 0
for
i = 0
or
i.
results,
then,
guarantee
of an asymptotic solution U0 = 0
solution within
to
provided
satisfy
llz(0,0)l[ < ~
and
llz(l,0)il < ~ .
Indeed,
we then have
0 i l~z(t,a)lJ ~ m(t,)
where
satisfies
gin" = %(m,t,),
0 < t < i,
m(i,)
i = 0
and
i.
behavior
of
follows
results
of Howes
(1978)
c.
internal
transition
layers
problem
y" + f(y,t,c)y'
+ g(y,t,s)
= 0,
0 < t < i
where f2 (yl,Y2, t, ~) y = Y2 , f(y,t, ~) = 0 gl (yl,Y2) and g = Y2 This system decouples into the two nonlinear scalar equations
Y2
eY2 + Y2Y2  Y2
and
ey~ + fl(Yl,t,c)yl
+ [f2(Yl,Y2,t,E)y~
+ gl(yl,Y2,t,e)]
= O.
147
If Howes
and
it follows from
UL( ~
 i) = 0,
UL(0) = Y2(O)
on
0 < t < t* = ~
(i  Y2(1)  Y2(0))
UR(U ~  i) = O,
UR(1) = Y2(1)
on
t* < t ! i,
i.e.,
f
Y2
U
uL(t) = t + Y2(0),
0 ! t < t* t* < t ~ i.
uR(t) = t + Y2(1)  I,
Thus,
t*
one elsewhere) becomes unbounded there. near t* from UL(t*) Y2(O) to and
For other
relations between the boundary values bilities occur (cf., e.g., Howes).
One must generally expect the transition layer at corresponding discontinuity let's assume that to the equation for there in = 0 YI"
t*
in
Y2
to generate a however,
f2(Yl,Y2,t,0) YI"
fl(VL,t,0)vL + gl(VL,U,t,0)
= 0,
0 < t < i,
VL(0) = Yl(0)
and
= O,
0 < t < i,
vR(1)
= Yl(1).
Yl
will be provided by
vR(t)
> 0
rVR(0)
(vR(0)  Yl(0)) J q for q between vR(O) and (including) VL(t) on Yl(O). Similar conditions would imply fl (s'0'0)ds
>
148
near
t = i.
fl(VR(t),t,0) > 0
on
tR < t < i
while
fl(VL(t),t,0) < 0
on
0 < t < tL
with
tR < tL,
we can expect
Yl
J VL(t), Yl
V
0 ~ t < t t < t ~ 1
VR(t),
as
c + 0
in
(tR, t L)
such that
J(t) = O,
J'(t) ~ 0
for VR(t) r J(t) = J fl(s,t,0)ds VL(t) (cf. Howes (1978)). Pictorially, we will have limiting solutions Y2 and Yl as
Y2 uR
Y2 (i)
..............
Y2 (0) Figure 2
yl
A
149
vR
b t
t t* 1
Figure 3
Note that
Y2
has a jump at
and
' Y2
has a jump at
t*,
corresponding to a
Acknowledgment We wish to thank Warren Ferguson for his interest in this work and for calculating the solution to the first example.
References i. N. R. Amundson, "Nonlinear problems in chemical reactor theory," SIAMAMS Proceedings VIII (1974), 5984. E. A. Coddington and N. Levinson, "A boundary value problem for a nonlinear differential equation with a small parameter," Proc. Amer. Math. Soe. 3 (1952), 7381. D. S. Cohen, "Perturbation Theory," Lectures in Applied Mathematics 16 (1977) (American Math. Society), 61108. J. D. Cole, Perturbation Methods in Applied Mathematics, Ginn, Boston, 1968. F. W. Dorr, S. V. Parter, and L. F. Shampine, "Application of the maximum principle to singular perturbation problems," SIAM Review 15 (1973), 4388. A. Erd~lyi, "The integral equations of asymptotic theory," Asymptotic Solutions of Differential Equations and Their A~.!ications (C. Wilcox, editor), Academic Press, New York, 1964, 211229.
2.
3.
4. 5.
6.
150
7.
A. Erd~lyi, "Approximate solutions of a nonlinear boundary value problem," Arch. Rational Mech. Anal. 29 (1968), 117. A. Erd~lyi, "A case history in singular perturbations," International Conference on Differential Equations (H. A. Antosiewicz, editor), Academic Press, New York, 1975, 266286. W. E. Ferguson, Jr., A Singularly Perturbed Linear TwoPoint Boundary Value Problem, Ph.D. Dissertation, California Institute of Technology, Pasadena, 1975. P. C. Fife, "Semilinear elliptic boundary value problems with small parameters," Arch. Rational Mech. Anal. 52 (1973), 205232. P. C. Fife, "Boundary and interior transition layer phenomena for pairs of secondorder differential equations," J. Math. Anal. A ~ . 54 (1976), 497521. W. A. Harris, Jr., "Singularly perturbed boundary value problems revisited," Lecture Notes in Math. 312 (SpringerVerlag), 1973, 5464. F. A. Howes, "Singular perturbations and differential inequalities," Memoirs Amer Math. Soc. 168 (1976). F. A. Howes, "An improved boundary layer estimate for a singularly perturbed initial value problem," unpublished manuscript, 1977. F. A. Howes, "Boundary and interior layer interactions in nonlinear singular perturbation theory," Memoirs Amer. Math. Soc. F. A. Howes, "Modified HaberLevinson crossings," Trans. Amer. Math. Soc.
8.
9.
i0.
ii.
12.
13.
14.
15.
16. 17.
W. G. Kelley, "A nonlinear singular perturbation problem for second order systems," SIAM J. Math. Anal. M. Nagumo, "Uber die Differentialgleichung Math. Soc. Japan 19 (1937), 861866. y" = f(x,y,y')," Proc. Phys.
18.
19.
R. E. O'Malley, Jr., Introduction to Singular Perturbations, Academic Press, New York, 1974. R. E. O'Malley, Jr., "Phaseplane solutions to some singular perturbation problems," J. Math. Anal. Appl. 54 (1976), 449466. P. R. Sethna and M. B. Balachandra, "On nonlinear gyroscopic systems," ~echanics Today 3 (1976), 191242. W. Wasow, "Singular perturbation of boundary value problems for nonlinear differential equations of the second order," Co_~. Pure Appl. Math_. 9 (1956), 93113. W. Wasow, Asymptotic Expansions for Ordinary Differential Equations, WileyInterscience, New York, 1965 (Reprinted: Kreiger, Huntington, 1976). J. Yarmish, "Newton's method techniques for singular perturbations," S I ~ Math. Anal. 6 (1975), 661680. J.
20.
21.
22.
23.
24.
CONDITIONS
H.W.Knobloch
1.1ntroduction The lecture nonlinear a dynamical (1.1) is intended systems to give a survey on some recent research of "system" refers to in
theory.
differential
equation
u~U,
U being
an arbitrary
set
(control region).
u=(ul,...,um) T
are finite
We admit specializations of t
C~functions
control
functions).
A pair and
where
control function
of the differen
of (1.1) . so called
of (1.1),the
We present
a unified
approach
to both problems
compared with other relevant work in this results and requires roughly rather ele
more accurate
mentary
The background
work consists,
speaking,
essentially
in a careful
analysis
152
eq.
This analysis
coming paper. 2. Local controllability We consider tion) a fixed and cones of attainability. of (1.1) (reference solu
solution [0,t I]
(u(.),x())
on some interval
. The system is said to be locally solution at the terminal point tI into any point of initiaof the The
controllable x I = x(t I )
trajectories
x o = x(0).
This concept
controllability to sufficiency
theory.
of attainability.
approximations from
of all points
attainable
guments
theory,
Maximum Principle
involves
of a cone of attainability.
since conex
cone does not yield the best possible have been made to find suitable most attention
and attempts
so far is the of a a
Principle".
outline
a derived
which contains
cone ond the cones used in the work of Krener. essentially of two steps. SteD I. We pick a
consists
point
(u(~),x(~))
153
state space.
tailed description
coming paper). We collect all ndimensional vectors which appear as first nonvanishing coefficient in any formal power series which can be generated in the following way. Consider an admissible control function u(,k) which depends upon some positive parameter u(t,k)=u(t) (= reference control) ~0. k and
except a neighx(',k) be
Let then
eq. (1.1)
(with
t=0 the same initial value as the reference trajectory. k=0 of x(t,k)x(~).
lit, 0 ~ t ~ tl,
and finally the convex cone generated by its elements is taken. The result is then the cone of attainability which will be denoted by K and which provides the basis for the subsequent considerations. Theorem 2.1 Hestenes, terior to cf. ~. K is a derived cone for ~ a t [3]). In particular, if x(t I) (in the sense of then x(t 1) is in
K = ~n
If the reference solution is optimal then by standard arguments (state augmentation technique and application of the generalized [3]) one can derive from Theorem 2.1 first These conditions can also be obtained
in a more geometric way via the following theorem. M will denote a subset of the state space which is defined in terms of equations and inequalities. The notions "regular point of M, relative interior
point of M, tangent cone T at some regular point of M" are used in the sense of [4], Chapter VII, p. 320.
154
Theorem 2.2
Let
be a subset of the
Rn
and let
x(tl) of x(tl) to
regular point of
M. If there exists
a neighborhood is relative
does not contain a point of ~ w h i c h K R n) and the tangent cone . This implies y() T to M at
interior
x(tl)
the existence
of a nontrivial relations
state
variable
(2.1)
and all .
tE[0,t I]
applications
to optimality
both in equality
and inequality
where
3. Singular
extremals.
General
is optimal
of a singular application
extremal
of the Pontryagin
frequently
turns
there is particular
interest
in the Pontryagin
conditions
order".
In the lecture
order conditions in the two basic The second depicts IIt. It gives rise cases of which are to Robbins and
special
in applications
attributed
In case of a multivariable
155
which seems not to be known so far (cf. related work by Kelley, Moyer, Jaeobson, Krener and others). conditions is not just a theory, but
Kopp,
The theory
straightforward
on some nontrivial
out these
interest
equations
Firstly we introduce
bility matrix into the nonlinear tonian system for the state system
(3.1) Let us
Let us consider
that is the
which arises
x = f(x,u) assume for
= y T . f(x,u):
simplicity
that
is
scalar.
Take
the
formal
vth
time derivative
(~H/~u)(x,y,u)
= yTfu(X,U)
(3.1).
yT'b
It
is easy to see that it can be represented where b is a ndimensional x,u and further system with the
certain
/ %
functions of
(i.e. vth
It now turns out that in the general nonlinear quence of the b provided they are defined variables
as above,
namely
functions
of the independent
~,~,...
role not only for the first order necessary not surprising) conditions.
conditions
The second idea  which so far seems to represent of our approach  is to study systematically
invariance
156
(3.2)
In other words,
u ~ u(x,v)
we consider along w i t h the given system all those depend upon the state. elaborated It turns
the control
to the sets
this is somehow
to other relevant work in this field  to avoid the usage of Liealgebratheory. ~ . many independent a mdimensional the sequence variables The symbol
vector.
lUo,Ul,... I ;
and finitely
will be denoted by
g(x, U
) for shortness.
are infinitely If g
with respect
[f,g] = gxffxg
f=f(X,Uo) (1.1)
is the right hand side of the given differen(with uo instead of u). Hence
tial equation
(4.1)
r(g) = [f,g] + ~
i=O
(~g/~ui)'ui. I
g ~ F(g) represents vectors a linear g=g(x, U b ).
The mapping
which
bo=fu(X,U o) of
the operator
etc. instead
u o , Ul,U 2 ....
157
We now turn to the case of a multivariable of the b is then played by a sequence column vectors of u
control
(4.2)
~o = (~fl~u~)(X,Uo)
~ are
'
= r v ( ~ o) , v=1,2, . . . .
of of x, U ui = IUo,Ul,..l for and
The elements of
C~ functions
in the components B
1 ~ i ~ v As
also can be obtained with the help of the = yT.f(x,u) and its Jacobian matrix
H(x,y,u)
= yTfu(X,Uo)
is differentiated
is carried out according to the rules , y= fx(X,uo)Ty , ui = ui+1' i=0,I .... yTB (x,u) , i.e. we have
in the form ) .
By(x, U
This remark leads to a further interpretation Let us take the (1.1), (4.5) i.e. B along a given reference
Bv of
u(.),x(')
By(t) = Bv(x(t),
It is then easy to see that the following relation holds dv (yTf. (x(t),u(t)) = y ~ v ( t ) dt v if the differentiation of y is performed (4.7) y =  fx(X(t),u(t))TY of this relation and the Pontryagin Maximum order necessary conditions": y(') an adjoint state(4.6)
An obvious consequence
vector
(i.e. a nontrivial
158
columns
of
By(t),
v=0,1 . . . .
more general
x('),
u(t)Eint
has to be made
Theorem 4.1
the linear
Bv(t),v=0,1, . . . .
the
(4.7) U(cf.
formally
This
one to express
order,
An important
is the following
The following
relation
holds identically
in
x, U
for
operations
differentiation forming of
Liebrackets
A proof of Theorem 4.2 which is based on the invariance mentioned paper. in connection with
principles
159
5. The generalized Given a reference us assume that ClebschLegendre solution condition. on some interval I and let
u(.), x(')
u(t)EintU Rn
for all tEI. We denote as before by ~ ( t ) which is spanned by the columns of the
Let there be
(t)for
for some
tel
if
~ =,
(1)~/~
i,j=l
~
adjoint y(t) is
It follows then from Theorem 4.1 that for every n1 tel. If ~ ( t ) (i.e.
to ~ ( t )
if the multiplier
then c o n v e r s e l y ~ t )
simple form. is constant and equal ~I' .... '~m to the first
nonvanishing m
(5.1)
~
i,j=1
~i~Y(t)T(bB~/bU~J))(x(t),
~
U(t)),~=0,1,2, . . . .
~1,...~m ) and
160
As one observes from (4.3) and (4.4) the coefficient of also be interpreted as the quantity (5.2) ~ # d~ ~
Ci~j
can
x=x(t),y=y(t),
more close to what is known as generalized ClebschLegendre condition in the literature. Note however that the standard version of this condition in case of a multivariable controlconGerns the quadratic form (in indeterminates zl,z2,...Zm) i zizoh~'J(x(t)'y(t)' i,0 as such, whereas each U(t))
our corollary yields the analogous statement for value of this quadratic form.
i n d i v i d u a i
We state a further result which is an immediate consequence of Theorem 5.1. In the following corollary the reference solution is not required to be optimal and the rank of mal. Corollary. 2. (5.3) for Let ~ > 0 be an integer such that (bBJ/bu(i))(x(t), ~(t))E ~ ( t ) v=0,...,~1 tel, Assume furthermore i,j if V=~ ~(t) need not to be maxi
n o t
and all
i,j=i,...,m .
161 6. Higher order equalitytype These are conditions conditions. y(t)Ta = 0 and they arise, An
of the form
A aE I~ . All elements t
In particular
it need not
~ _> 0
be an integer such that the following ele~(t), for for every tEI:
U(t))
v ~ ~, i,j=l ..... m
for
i,j = I .... m.
The elements ~ (~B~/bu~J))(x(t), "If t for every tEI U(t)) and i,j=1,...,m.
There is an important special case of Theorem 6.1 which is known (or rather its consequences for singular extremals are known).
Corollary Then
1.
Let
f(x,u)
be a linear function in
~t
Proof.
If
is linear in
u, then
Bo = fu
is independent
from
162
U and hence
b o/bU o
Bi, {j)
is zero identically in
x,
. It follows of Theorem
then from Corollary 2 to Theorem 5.1 that the hypotheses 6.1 are all satisfied if one takes ~=I .
One observes that the conclusion of Theorem 6.1 can also be phrased in this way: The convex cone generated by the elements of ~ t tains the linear space generated by the union of ~ ( t ) ments con
in the first corollary of Theorem 5.1. Since every linear subspace of "lit is orthogonal to the multiplier II t y(t),~(t) is necessarily n1. Hence
if it has dimension
and a straightforward
induction argument leads us to the following result. Corollary 2. Assume that ~(t) is the maximal linear subspace con~It , for tEI, for
tEI. Assume furthermore that (5.3) holds for all and for v=O,...,a ,~
7 ApPlication to sensitivity analysis. We wish to touch briefly upon a further application of the foregoing results which underlines a certain advantage of our approach. the standard techniques of general properties the cone K Since
of derived cones only, they can be applied to in Sec. 2  the same cone from
which we have deduced all necessary conditions discussed in this lecture. Thereby one arrives on sensitivity results which take the
163
effects
into account.
Sensitivity
is concerned
with estimates
for the
of the
account
can be found in
[5] where also an illustrative Let us consider fold consists direction paramter. value an optimal
is discussed.
control problem where the terminal manix 1. We now change xI by x1+kp, xI in a certain
ef a single point
function V'
side V'
derivative
can be estimated linear functional those multipliers Maximum Principle if we let satisfy y
from above in terms of the values which a certain assumes y on the (suitably normalized) set of all
holds
which
a restriction
also a lowering
cone is contained
References.
in nonlinear
systems, Academic
Systems,
and L. Cesari
eds.,
1977,
pp. 157174. The high order maximal principle extremals, SIAM J. Control and its application 15 (1977)
Optimization
164
[3] M.R. HESTENES, Calculus of Variations and Optimal Control Theory, Wiley, New York 1966 [4] H.W.KNOBLOCH und F.KAPPEL, Gew~hnliche Differentialgleichungen, B.G.Teubner, Stuttgart, 1974. [5] B.GOLLAN, Sensitivity results in optimization with application to optimal control problems. To appear in: Proceedings of the Third Kingston Conference on Differential Games and Control Theory 1978.
I n s t i t u t Meth~matique Universit~ de Louvain B134B LouvainlaNeuve Belgium I . INTRODUCTION Much work has been devoted in recent years to the s o l v a b i l i t y o f nonlinear equations o f the form
(1.1) Lx  Nx = 0
in a Bsnach space, or to the study of the range of L  N, when L is a Fredholm mapping of index zero and N satisfies some compactness a s s u ~ t i o n . graphs[ 8 ] , [ 1 1 ]
and[13] . Basic for this study is the
reduction
(u,v) = (u + JQ~u+v), KpQN(u+v)) where P and Q are continuous projectors such that (1.3) Im P = ker L , Im L = ker Q ,
KpQ i s the associated generalized inverse o f L and J : Im Q ~ ker L i s an isomorphism. The compactness assumption on N generally implies that (JQ + KpQ)N i s a compact mapping on some bounded subset o f X and, P being by d e f i n i t i o n of f i n i t e rank, (1.2) i s a fixed
point problem for a compact operator in X and degree theory is available in one form or another. If one replaces the Fredholm character of the linear mapping L by the mere projectors P and Q satisfying (1.3), then P is no more compact
existence of continuous
, chapter 13).
verifies some growth condition, and L belongs to some class of linear mappings havif~g in particular compact generalized inverses. satisfied for the abstract formulation semilinear wave equations. Those assumptions are in particular
mappings is rather long and uses a combination of the theory of maximal monotone operators, 5cheuder's fixed point theorem and a perturbation argument.
166
In this paper, which is the line of the recent work of one of the authors ~B]for the case of a Fredholm mapping L, we consider problems with dim ker L non finite
by an approach which is closer in spirit to the continuation method of Leray and Schauder [14], although we still have to combine it with other powerful tools of
nonlinear functional analysis like the theory of Hammerstein equations and of maximal monotone operators. We first obtain a continuation theorem for Hammerstein equations
(Section 3) whose proof requires an extension of some results of De Figueiredo and Gupta ~ g i v e n in Section 2. This continuation theorem is applied in Section 4 (1.1) in a Hilbert space under regularity and with the
assumptions slightly more general than the ones of Br@zisNirenberg growth restrictions replaced by a condition of LeraySchauder's of some set.
This existence theorem is then applied in Section 5 to abstract problems type, the first one corresponding to a result of Cesari and Kannan
of LandesmanLazer [9]
, and the second one being essentially a version of the In Section 6, the main theorem of Section 4 is applied
result of Br~zisNirenberg.
partial answer to the question raised in 6 0 ] about the solvability of this periodic problem for differential equations in infinite dimensional spaces. we consider equation In Section 7,
type theorem when the assumptions of monotonicity and compactness are replaced by some assumptions of strong continuity. Another problem in differential equations and leading to equations of the type (1.1) with an infinite dimensional kernel is the periodic boundary value problem for first order ordinary differential equations of the form x' = f(t,x) in an infinite dimensional Banach space. in [5] and Z 6 ] a n d Such a problem was considered by Browder The interested reader can consult the leading in
recent paper ~ 9 l w h e r e some of the results of Browder are generalized, particular to existence theorems of the LandesmanLazer's
167
A pair (M,N) of mappings from H into H is said to be Hammersteina an___.dd b if the following conditions
0 ~ b ~ a . N is demicontinuous,
(V u ~ H ) ( V
(V u ~ H ) ( V
v~H)
v ~H)
i.e. if
vl 2 _~ (Nu
Nv,u 
v)
This class of pairs of mappings is related to the unique solvability of the abstract Hammerstein equation
(2.1) x + MNx = f
for every f ~ H, by the following results, which generalizes and completes a theorem of De Figueiredo and Gupta [ 1 0 ] PROPOS3TION 2.1. constants .
Let (M,N) be. a.. pair .of hcompatible mappinqs from H t__ooH, with Then.~ for every f ~ H, equ..ation (2.1) has a unique sol.u.tion.
a an__.d.d b
If, moreover ,
(2.2) M(O) = 0 ,
(2.1) s a t i s f i e s Ix  f  ~
(a  b )  I I N f l
We only sketch the proof, details of which will be given in a subsequent (2.1) is trivially equivalent to equation
y + MN(y + f ) = 0 ,
Equation
and hence to
(2.4) if O~Ty,
T : H .~ 2 H i s d e f i n e d by
Ty =  M  1 (  y ) + N ( f + y ) .
By the assumption on M and N and basic results of the theory of maximal monotone operators (see for example [ I] ), it is easy to show that T is maximal monotone To
168
Under some more assumptions, one can obtain results about the continuous dependence of the solution of equation (2.1) with respect to M, N and f . The formulation of the following Proposition 2.2 is modelled after a theorem of Brezis and Browder (~2] PROPOSITION 2.2. , Proposition 5) and the proof will be given elsewhere. H, ( M n , N n ) n ~ N .
be a sequence of pairs of hcompatible mappinqs with .c.onstants a and b , and (fn)n~_N . be e sequence of elements of H which converqes to f. conditions hold : (I}
(2) (3)
(4)
U N (5) is bounded . n n ~ N* F o r each u ~ H, M N u   ~ MNu i f nP=o . n For each uC. H, N u   ~ N u if n~=~. n For each n ~ N*, Mn(Oi = 0
u = (I + MN)If , u = (I + M N )  I f n n n , (n E N * ) ,
5 C H,
Then, if n
If we call a mapping
F : H PH
bounded subsets of H, we have the following immediate consequence of Proposition 2.2. COROLLARY 2.1. H~H bounded. Let (M,N) be a pair of hcompatible mapp.~nos, with M(O) = D s n d N :
Then (I + MN) I H ~H is continuous. closed when F{S) is closed for every closed
subset S of H, we deduce at once from Corollary 2.1 the following COROLLARY 2.2.
a
I + MN : Hp H
i~s
cl,psed mappinq.
3. A CONTINUATION THEOREM OF LER.A..Y.SCHAUDER' TYPE ~ Let still H be a real Hilbert space with inner product ( , ) and corresponding norm I. , let I = [0,1] ~Q and, if E C_H and T : E x I ~H , (x,~) wpT(x,~), . fr ~'~ ,
I, by T ~
the m a p p i n g
C : cl~.
x I ~ H .
The following result is a continuation theorem of Lerayichauder's type for mappings which are not necessarily compact perturbations of the identity. H Recall that F : EP
169
THEOREM 3.1.
Assume that the mappinqs M, N and C satisfy the followinq conditions. a enid b such that, for each ~ _ I, the pair ( M , N~)
with constants a an_..~db , and M~(O) = 0 is bounded . and, for each x ~_H,
for each ~ I .
the mappinq
p ~. N(x, p )
i s continuous.
3. For .each x~ H and each V ~ I, the mappinq
~ ~~ M(N(x,y ) , ~ )
is continuous.
4. 0 ~
(I + MoNo)(D~)
x I ~
.
and CO = 0 .
5. C : c l ~
H is compsct
f o r every ( x , ~ )
Then, for each ~I,
fr~.
x I .
equation
X + MpN~x + C x
has at least one solution Proof. Let us define, x ~ ~L . for each ~
I, the mapping
+ MpNp .
T~:
H ~H
by
TM =
It follows therefore 2.2 thst the mapping
Consequently, ~.x I]
i s compact on c l U
Moreover, i t
cI(TF(~L)) c
T (cl ~
) .
Therefore, T~ being moreover onetoone, one has f r T (~) = c l ( T B ( ~ ) ) k i n t ( T M ( ~ ) ) = cI(TM(3"L)) ~ T (~L) C C T~(cl~) ~ T~.(J'L) = T (fr~) , f r U, one has
170
As, by assumption (5), K0 = 0 , all the conditions are satisfied to apply the usual LeraySchauder's continuation theorem [14] to the family of equations
(3.2)
implying that, for each
y + K(y,~)
= 0
~ 6
I,
(y,p)
and hence, l e t t i n g
y = Tpx
U,
xE~C
o f U, we see t h a t
satisfie
der's type continuation theorems due to Browder [ 73 and Br@zis and Browder
[2].
WIT H
4. A CONTINUATION THEORE M FOR SOME NONLINEAR PERTURBATIONS OF LINEAR MAPPINGS AN INFINITE DIMENSIONAL KERNEL
Let still H be a real Hilbert space with inner product ( ~ ) and corresponding norm form I'~ We shall be interested in existence results for equations in H of the
(4.1)
where
Lx = Nx L is a linear and N a not necessarily linear mapping from H to H satisfying Following the line of recent , and in
some regularity assumptions we shall describe now. work by Mawhin [16], Br~zis and Haraux ~ ]
contrast with most of the literature devoted to equations of type (4.1) (see for example the s u r v e y s t B ] dimension. More precisely, let dense domain (4.2) dom L L : dom L ~ H ~PH Im L be a closed linear operator with such that , D I ] , D 3 ] ) we shall not assume t h a t ker L is of f i n i t e
Im L = (ker L)"L
Consequently, using the closed mapping theorem, the restriction of L to dom L ~ Im L is a onetoone linear mapping onto Im L with a continuous inverse
K : Im LP dom L ~ Im L . We shall denote by P the orthogonal projector in H onto the (closed) subspace ker L , so that condition (4.2) is equivalent to Im L = ker P . Let now N : H~ H Nx n be a bounded mapping which is moreover demicontinuous on H, for every sequence (Xn) n ~ N* which converges to x in H.
~ Nx
171
H and every y ~ H,
0 .
Let us assume that the mappinqs L, N, K, P defined abpve verify the : is monotone. ~).C H such that 0 ~ ~ and
fol.lowin.q F onditions
I. I  P  N : H~ H
Lx + (I  ~)Px  ~ N x
(4.3)
has at least one solution
Lx =  ( 1
x ~
 ).)Px +~Nx
By a result o f ~ 5 3 (see a l s o ' t 3 3 ),
dom L (~ ~L
(4.4)
and hence, by assumption
(4.5)
for every in dom L . (x,~) ~ Let ~
x +~P(P  N)x  ~ K ( I
 P)Nx ~ 0
fr ~. x ]0,1[ , because the solutions of (4.4) are necessarily ~0,I~ be fixed; then, one immediately obtains , using the (I), for every x G H and y E H, and each ~ ( I ,
X1]%px _ %py]2 ,
 P)Nx
0, ands, by ( 4 , 5 ) ,
x + ~.'.XP(P  N ) x  f f " ~ K ( I
for avery ( x , ~ ) ~ fr ~ x I, because, when
 P)Nx
~=
#~ 0
being bounded, it follows from the first step, the compactness of and the properties of L that one can find sequences (~ ) n n~N* '
P)N on c l ~ ~ .
(~n)n~.N . , (4.6)
auch t h a t
)~n.~ 1
if
n~oo, X h ~ ~ 0 ~
n
, n~N~,
n n
Lx
= (I
 Xn)PXn + ~ Wx
n ~N*,
172
and, if we write
Yn = (I P)x n , z n = Px n ,
y~ ,
and z E Zn~ z
(4.6),
(4.8)
Using
(Yn  A n 1(1 
(4.7), we can go to the limit in (4.8), which gives (y  Ly  {I  P  N)v, x  v) ~ 0 this implies
for every v ~
H.
and hence Lx = Ly = Nx and the proof is complete. REMARK 4.1. It is immediately checked that L has the same properties as L, with Consequently, the assumption I  P  N is
is monotone
if (2.ii) is replaced simultanuously by 2.ii'. Lx  (I  ~ ) P x  % N x ~ 0 for all (x, ~ ) ~ (dam L ~ fr~L) x ]O,i[ .
I  P  N
We shall show in this section how Theorem 4.1 allows simple proofs for abstract LandesmanLazer problems, i.e. results about the range of L  N when N satisfies The first theorem is modelled after a result of Cesari This assumption I  P  N or I  P + N.
and Kannan Z 9] for the case where ker L is finitedimensional. is suppressed here at @ e exp.nse of a m o n o t o n i c i t y condition on C O R O L L A R Y 5.1,
listed at the beoinninq of Section 4 an d that the followinq assumptions hold; a. I  P  N : H~H b,' K(I  P)N : H~H K(I  P ) N ( B ) ~ s c. There exists r~ is moptone. is completel.y continuous, i.e. continuous and such that
relatively compact for every bounded subset B o f H. 0 such that, for all x ~
iK(I e)Nx~ r
H,
.
173
H for which ,
IPxt = R and
one has
(I  P)x~ ~ r
(Nx,Px)
0 .
Then t equation (4.1) has a t l e a s t one s o l u t i o n . Proof. We shall apply Theorem 4.1 with the open bounded s u b s e t ~
Q =
of H defined by
~xEH
IPx~R
and
(I
P)x~<r},
and and
l(I
 P)xl_~r~
S2 = {X ~ H : IPxl ~ R
l ( I  P)xI= r }
~]0,I[, equation
(5.3)
i.e. such that
l(I  P)xI = ~ K ( I
x ~ 52. Now, (I
 P ) N x ~ Xr
<
r , ,
because, by the o r t h o g o n a l i t y o f P, (PNx,Px) = (Nx,Px) = (PNx,x). assumption (d) and (5.3), one has x ~ 5I. Thus, assumption(2)
satisfied with our choice o f  ~ REMARK 5.1. conditions replaced by a'. I  P + N : H ~ H d'. (Nx,Px) ~ 0 By using
Remark 4.1, one easily obtains that, in Corollary 5.1, (d) can be simultanuously and respectively
is monotone.
w i t h the same conclusion f o r the equation ( 4 . 1 ) . We shall now deduce in a very simple way, from Theorem 4.1, a recent result of Br~zis and Nirenberg[4] . Assume that L : dom L C H ~H satisfies the conditions listed at x E dom L,
174
(5.4)
for all x ~ dom L.
(Lx,x) ~
Let now B : H  ~ H
 ~ILx~2 ,
We have
the following result. COROLLARY 5.2. Assume that the mappinqs L a n d s satisfy the above conditions and
that the followinq conditions hold; a. I  P + B : H ~ H ~s monotone. ..i..s...complete.l.E..continuous. O f ~ ~, such th..a.t,, for all x ~ H and y ~ H,
, with
(5.5)
Im(L + B) .
f ~ i n t ( I m L + cony Im B)
and let us apply Theorem 4.1 with N = f  B. By the assumptions of regularity made
on L and B, it suffices to prove that the set of solutions of the family of equations
(5.6)
Lx + (I  X ) P x + k B x = %f
, because then condition (2) of Theorem
t . Bw I l
( 1 ~ i ~ _ n ) and
n ~ t i = I. i=1
It
f o l l o w s then from
(5.6) that
(1 ~)Px + ~(Bx 
n
~ t.Bw,)
i=1
* ~h
= ~Lv
Lx
Taking the inner product of this equality with x and using (5.4) and (5.5), we obtain
(1  X ) l P x l z + A ~ t i ( ~ r  l l B x l 2  c(w.))~. X ( h , x ) i=1
~
and hence
~ X(Lv,x) + e  l l L x I 2 ~
~lLv
 ( I  P)x~ +
@llLxl 2 ,
(5.7)
But, by (5.6) and
~1~Bxl 2 
~ i=I
t.c(w.) i i
+ ( h,x)~Lv
l ( I  P)x{ +
(~@)IILx~2.
(5.1) with
which, together with (5.7), gives, for all h E H of sufficiently small norm,
(h,x)
(8I
175
As
c"'(h)
@I ~  I
Ixl~
which achieves the proof. REMARK 5.2.
R,
More general versions have been given by Br~zis and Nirenberg in [4] We have restricted to this one for simplicity.
Let us also notice that Br~zis and Nirenberg assume that B is monotone instead of
I P + B.
(a) is replace d b ~
Proof.
In particular,
~Bxl*~
if
Ix~~,w~
We shall refer to [ 4 ] for an interesting application of Corollary 5.3 to the existence of generalized solutions in L 2, 2Ttperiodic
wave equation
utt  Uxx
f(t,x,u)
A generalization of some of the results of [17]is obtained which replaces some Lipschitz condition in u for f by a monotonicity assumption. 6. AN APPLICATION TO A SECOND ORDER PERIODIC BOUNDARY VALUE PROBLEM. IN A HILBERT SPACE In this Section we shall apply Theorem 4.1 to the second order periodic boundary value problem
(6.1)
x"(t)
x(O) 
= f(t,x(t)),
x(1) = x'(O) 
t ~
x'(1)
I =
[0,1] ,
,
= 0
f : I x HI~ HI, with HI a real Hilbert space with inner '~ I " One could consider similarly the
= o = x'(1)
which share with the periodic ones the feature of furnishing a noninvertible part in the abstract formulation of (6.12).
linear
When the boundary conditions make this rather general results for
the solvability of the corresponding boundary value are available, even for HI a
176
dimensional, the general case with periodic or Neumann boundary cnnditions is much more difficult and ~ e s not seem to have been explored. results we obtain here may be of interest. Let H = L2(I,HI) with the inner product Hence, even the very special
(u,v) =
and the corresponding norm
(Nx)Cs)
JI
(u(s),v(s))ds
for
x ~
H.
subsequent paper, we shall directly make our regularity assumptions on N instead of on the original mapping f . Let us assume that :
I
for all x ~ H and y ~ H . dom L ~ H by
x(O)  x ( 1 )
let
L : dom L C
H ~H~ x~Px", so
Im L = ~x ~ H
~I x(s)ds = 0 } =
(ker L) "L
ker L = Im P
with P : H   ~ H
Px = ~ I
x(s)ds
THEOREM 6.1. R~
Assume that the condit~pns above hold for N and that there exists and all x ~ HI with ~ x  1 ~ R, one has
(6.3)
(f(t,,x),x) I
(r/21) 2 .
Th.en~ problem (6.12) has a t l e a s t a (~aretheodor.v) s o l u t i o n . .Proof. We s h a l l apply the v a r i a n t o f Theorem 4.1 mentioned i n Remark 4.1 to the
177
Lx = Nx . C l e a r l y , the sssumptions we have made imply t h a t I  P + N i s monotone and K(I  P)N
completely continuous
K is a
(6.5)
and hence,
Ix"l = ILxI(INx
O = (I  %)Px + APNx ,
(r
Letting
x = u + v, with
_~ ItrI((21~)23 I/2)
= r'
lu~ = u  1
~ 
I,
Ix(t)~l~lUll
max t~I
l v ( t ) l 1 ~, R ,
But (6.5) impliesp after having taken the inner product with x,
O = (I  ~ ) l u  2 + %(PNx,x) , i.e.
0 = (1  ~ ) l u  2 + ~ ( N x , x )
 ~(Nx,v)
0 >~ (1  ~ )  u 1 2 + a contradiction.
A(r/2~) 2 ~(r/(2It)
2)
= (1  % ) l u l
2 ~
(1  ~)(R + r ' ) 2 ,
Therefore
lul 1 < R + r'
and hence, by ( 6 . 6 ) , Ixl 1 ~ lul 1 + Iv~ 1 and the proof is complete. < R + r' + ( 2 T c )  2 r
178
7. A CONTINUATION THEOREM FOR SOME NONLINEAR PERTURBATION OF LINEAR MAPPINGS IN REFLEXIVE BANACH SPACES Let X be a real reflexive Banach space, Z a real normed space, L : dom L ~ X ~ Z Q : Z~Z
a linear mapping such that there exist continuous projectors P : X  ~ X , for which ker L = Im P, Im L = ker Q ~ ~ P  = ~ and such that there exists a linear homeomorphism denote by Kp,q : Z ~ dom L ~ ker P J : Im Q~ker L .
Let us
dom L ~ ker P is the inverse of the onetoone and onto restriction of L to dom L ~ ker P . let N : c l ~ L ~ Let I~LC X be a bounded open convex subset with O~L, and
Z be a (not necessarily linear) mapping such that the mapping (dq + KpQ)N : c l ~ l  C X ~ X
p
is stronqly continuous on c l ~ 
(JQ + Kpq)N(x)
for X a reflexive ~anach space, but the converse is not true (see e.g. L 1 2 ~ THEOREM 7.1. (7.1) for every (x, % ) ~ (7.2) has at least one solution. Assume that L and N satisf~ the conditions above and that Lx ~ (I  ~)JIpx + ~ N x (dom L ~ fr &"L) x ] 0 , 1 ~ . Lx = Nx Then equation
(7.4)
For each
x  ~ P x = ~ ( J Q + Kp,Q)Nx ,
~ ~ I, the mapping I ~P
~ ~ I = [0,I~ .
with ( I ~XP)  I = (I  ~ )  I P + ( I  P) ,
and hence the family of equations (7.4) is equivalent to the family of equations
179
x = (I  h ~ )  1 ~ J Q N x + ~KpQNx = T A ( x , ~ ) , in cl~rL .
i , is
x  T~(x,/~) ~ 0
fr~.x
~=
t and hence equation (7.3) has at least one ~]D,I[ . Let now (~n)n ~ N* be e a sequence in dom L ~ such
and (Xn) n ~ N .
 X )JIpx + ~ Nx n n n n
nEN*.
(7.5)
If we w r i t e x
n
Xn " ~nnPx = = Yn + z
n
~ n (JQ + KpQ)NXn
Yn = PXn ' n ~
,
N*,
n ~ N* .
then (Yn)n~N. and ( Z n ) n ~ N . are
with
bounded and, hence, the compactness of (JQ + KpQ)N and the reflexivity of X imply that, going is necessary to a subsequence, PXn = Yn .~b y ~ so t h a t , by (7.5),
z n ~ z Eker P if n~~.
one has
Consequently,cl~being
weakly closed, x ,
~ y + Z = x i f n~o~ , and x ~ c l ~ n and then, by the strong c o n t i n u i t y o f (JQ + Kp~)N, one gets (JQ + KpQ)N(xn) ~(JQ + KpQ)N(y + z) The uniqueness of the limit implies that
z = (JQ + K p Q ) N ( y + z ) , i.e. x  Px = (JQ + KpQ)Nx
J
It is easy to check that Theorem 7.1 can be used instead of Theorem 4.1
to prove the existence of a solution for the periodic boundary value problem (6.12) is assumptions I and 3 on N are replaced by the assumption I'. N maps H into itself and is strongly continuous on every bounded subset of H , the proof being entirely similar to that of Theorem 6.1. It is an open problem to us
to know if the result still holds if one replaces in (I') the strong continuity by the complete continuity without the monotonicity condition used in 5ection 6.
180
RFRN E EE E CS
I. H. BREZIS, "Op~rateurs maximaux monotones et semigroupes de contractions dens les espaces de Hilbert", Mathematics Studies 5, NorthHolland, Amsterdam, 1973. 2. H. BREZIS and F.E. BROWDER, Nonlinear integral equations and systems of Hammerstein type, Advances in Math. IB (1975) 115147. 3. H. BREZI5 et A. HARAUX, Image d'une somme d'op~rateurs monotones et applications,
I s r a e l d. Math. 23 (1976) 165IB6.
4. H. BREZI5 and L. NIRENBERG, Characterizations of the ranges of some nonlinear operators and applications to boundary value problems, Ann. Spuola Norm. Sup. Pisa, to appear. 5. F.E. BROWDER, Existence of periodic solutions for nonlinear equations of evolution, P~Rc Nat. Ace d. 5ci. U.S.A. 53 (1965) 11001103. 6. F.E. BROWDER, Periodic solutions of nonlinear equations of evolution in infinite dimensional spaces, in "Lectures in Differential Equations", vol. I, A.K. Aziz ed., Van Nostrand, New York, 1969, 7196. 7. F.E. BROWDER, "Nonlinear Operators and Nonlinear Equations of Evolution in Benach Spaces", Proc. S~mp. Pure Math., vol. XVII, part 2, Amer. Math. Soc., Providence, R.I., 1976. 8. L. CESARI, Functional analysis, nonlinear differential equations and the alternative method, in "Nonlinear Functional Analysis and uifferential Equations", L. Cesari, R. Kannan and J. 5chuur ed., M. Dekker, New York, 1976, 1197. 9. L. CESARI and R. KANNAN, An abstract existence theorem at resonance, Proc. Amer. Math. Soc. 63 (lg77) 221225. 10. D.G. BE FIGUEIREDO and C.P. GUPTA, Nonlinear integral equations of Hammerstein type with indefinite linear kernel in a Hilbert space, Indaq. Math. 34 (1972} 335344. 11. 5. FUCIK, "Ranges of Nonlinear Operators", 5 volumes, Universites Carolina Pragensis, Prague, 1977. 12. S. FUCIK, d. NECA5, J. SOUCEK, Vl. 50UCEK, "Spectral Analysis for Nonlinear Operators", Lecture Notes in Math. n 346, Springer, Berlin, 1973. 13. R.E. GAINES and J. MAWHIN, "Coincidence Degree and Nonlinear Differential Equations", Lecture Notes in Math. n ~ 568, Springer, Berlin, 1977. 14. J. LERAY et J. 5CHAUDER, Topologie et @quations fonctionmelles, Ann. Sci. Ec. Norm.
non lin~aires, in "M~langes Th. Vogel", B. Rybak, P. Janssens et M. dessel ed., Presses Univ. de Bruxelles, Bruxelles, 1978, 301315. 18. J. MAWHIN, LandesmanLazer's type problems for nonlinear equations, Confe~ t Sam. ~ t . Univ. Bari n= 147, 1977.
181
19. J. MAWHIN and M. WILEM, Periodic solutions of nonlinear differential equations in Hilbert spaces~ in "Proceed. Equsdiff 78",Firenze 1978, to appear. 20. K. 5CHMITT and R. THOMPSON, Boundary value problems for infinite systems of secondorder differential equations, J. Diffe!en~al Equations 18 (1975) 277295.
SOME CLASSES OF INTEGRAL AND INTEGRODIFFERENTIAL EQUATIONS OF CONVOLUTIONAL TYPE~)' E. Meister (TH Darmstadt
Abstract. S t a r t i n g w i t h c l a s s i c a l c o n v o l u t i o n a l i n t e g r a l equations on ~ n , t r a n s l a t i o n i n v a r i a n t operators and t h e i r symbol r e p r e s e n t a t i o n according to H~rmander are introduced. The various g e n e r a l i z a t i o n s concerning the domains G o f ~ n t e g r a t i o n lead to WienerHopf i n t e g r a l and i n t e g r o  d i f f e r e n t i a l equations on ~+ and on cones. Compound i n t e g r a l and i n t e g r o  d i f f e r e n t i a l equations of the p r i n c i p a l value and L1kernel type are discussed on ~n using r e s u l t s by Rakovsh~ik. Simonenko's theory of local type operators permits us to i n v e s t i g a t e g e n e r a l i z e d t ~ a n s l a t i o n i n v a r i a n t operators. WienerHopf i n t e g r a l equations w i t h s t r o n g l y s i n g u l a r kernels correspond to equations w i t h piecewise continuous symbols in both v a r i a b l e s . Convolutional equations on quadrants and wedges are studied via the theorey of operators of b i  l o c a l type. O. Notation. In the sequel the f o l l o w i n g a b b r e v i a t i n g terminology is used: ]Rn : = {x : x = (x I . . . . . Xn), x EIR} : n  dim. Euclidean space with
n
Z xvy v ~=1
infinite
element to each d i r e c t i o n
on
nsphere = {xE]Rn : I x l = I }
" = (~1 . . . . .
Un)eIN n : m u l t i  i n d e x
(o.5)
Ipl
D~
:=
~=1
~1
Cmo
~n
(0;6)
" = Dl
...'D n
where
(0.7)
: = i ~
; ~ = I, .... n
*)Extended version of a General Lecture at the Dundee Conference on D i f f e r e n t i a l Equations, 31st March 1978.
183
(0.8)
~P
= ~t
Pl
""'~n
~n
to
XE(X)
LP(o;G)
: characteristic
f u n c t i o n of set
E f such t h a t
(0.9) (0.10)
, or
, respectively,
G = ~Rn
ck(G) ck(~)
: space of ktimes c o n t i n u o u s l y on
:
differentiable
functions
space of f u n c t i o n s extendable to
G : = GU~G w i t h sup x~
(0.11)
II f l l ck(~ ) C~(IR n
max O~l~l~k
I D~ f(x)l
(Q.12)
= {fEck(IR n )
tr
= {mEC~(IR n )
: sup(l+Ixl2)k/21D'~m(x)l
Schwartz'sspace o f r a p i d l y
:
decreasing f u n c t i o n s dual to
F (0.13)
: ndim.
d e f i n e d by , ~E~ n
(Fm)(~) = 8 ( ~ )
: = ~ in

at ~ i
2J2~~n ~n
f ei<x,~>~(C)d~ by for f E ~ ' and onto i t s e l f , = ~(~) mE~, with property and
~'
(0.16)
(FD~m)(C)
f o r mE~
184
(0.17) (0.18)
<DUf,~>
GM where
= =
(l)lhI<f,DU~> {~EC~(~n)
for
: lDU~(x)l~Ipu(x)l ~ and
p~ is a polynomial depending on
"space of m u l t i p l i e r 3E , ~::
Functions of J~ and
m(m,9)
~(~,t~) ~(~,~)
~(~,~)
ac ~(Z,~) i f f (i) ~ ( A ) : = A(~C) is closed in (ii) (iii) ~(A) : = dim ~(A) = dim ker A < ~(a) : = dim ~/R(A)= dim coker A < ~(A) = ind A : = ~(A) ~(A) : "index of A". In case of 1. Introduction. and ~ = JC one w r i t e s ~(~) etc.
D e f i n i t i o n 1.1: AI,A2E~(~,~) are called "equivalent" i f f AIA2E~(~,V) Or: = {BC~(~,~) : B~A} gives the quotientspace with norm (1.1) llc~l : =
inf
II A+VII
:
There is the wellknown theorem by ATKINSON (1951) [2 ] Theorem 1.1 : E~(~,~) , iff BrE~, ~, (1.2)
Let ~ , ~ be Bspaces then AE~(~,~) is FredholmNoether, i . e~ there e x i s t a " l e f t  r e g u l a r i z e r " B~E~(~,~) ~ a " r i g h t  r e g u l a r i z e r " such t h a t
V I E ~ Z ) , and V 2 E ~ ) B~A = IZ + V1
and ABr = I~ + V2
t85
Remark 1.1 : This has been generalized to pairs of Fr~chet spaces ~,l~ and linear, densely defined, closed operators A (cf, eg, G H E G& KRE[N (1957) [42] OBR or P U S O F(1974) [71]. There you may find the properties of FredholmNoether RSD R operators listed. Let kELl(~n) , ~ELp (~n) , 1~p~ , then
(1.3)
and
k(xy)~(y)dy
e x i s t s a, e~ on
~n
(1.4)
Ii k~mllp
II kll
I"
II ~lIp
= I . F is then u n i t a r y on
[1~(~)]~(~)
~({)EL p' (m n )
the "symbol of A":
( c f . e@. TITCHMARSH [ 101]).Asolution e x i s t s  then uniquely  i f f (1.7) I t is given by ;(~) : ~(~) + ........ k(~) .~(~) ~a(~) : = Zk({) # 0 on ~n
(1.8)
1C(~)
: [1 + ~(~)] . f ( ~ )
A
where 1 + ~(C) is an element of the "Wieneralgebra" ~ : = { FLI(~n) the WienerLevi theorem. On the other hand let
(1.9) (p(D)@)(x) : S a .(D~m)(x) : f ( x ) e ~'
due to
be a d i f f e r e n t i a l
equation of order
This leads to the "problem of d i v i s i o n " which asks f o r conditions under which
(I.Ii)
~(~) = ~
E3 ~'
186 gives the transform of a s o l u t i o n . Taking of existence of the "fundamental s o l u t i o n s " f = ~c9' one is led to the question
(Thm)(x) : = m(xh)
is called a " t r a n s l a t i o n operator on ~ ". I f 3(1' 3( ~ is defined by < ~ f,m> : = <f, ~_hm> for fc~' and
me~ .
b) A e ~ ( 3 ( , ~ )
on
 is called " t r a n s l a t i o n i n v a r i a n t " i f (AThm)(x) = (ThAm)(x) for mc3E and hC~ n . proved in 1960 ~n _ or t h e i r , resp., AE ~(~,~).
(1.14)
Let ~ c 3 E , ~ c ~ ~
~A is called the "symbol of A". I f JL~E,~) denotes the set of a l l s~T~ibols to p l i e r f u n c t i o n s "  then i t is known that [51] J~(k2(~ n ) ) = L~(~ n) ~ ( L P ( ~ n) c L~(~ n) ~V~(LP(mn) , Lq(mn)) = {0} Extending to Frech~tspaces one has if l~q<p~ . AE~)  also called ' ~ u l t i 
or f o r i t s dual
~'
187
Now, e. g. on (1.16)
i n v a r i a n t equation
(am)(x) = (FloA.Fm)(x)
is found via Ftransformation: (1.17) ~a(~).~(~) = ~(~) e L2(~) . is " e l l i p t i c " , i. e. infI~A(~)l ~c~ n > 0
I t exists f o r every f (uniquely) iff A and hence [OA(E)]Ic L~(R n ) , given by (1.18) (1.19) ~(~) m(x) = = ~1(~).~(5) , thus
(Fl~AZ(~)'Ff)(x)
, i . e.
AI
: FIGA1.F e ~ ( L ~ n))
and C~(L2(~n)) too .
Remark 1.2 : For 1~p<2 this argument works only i f OAIEj{(LP(LRn)) which is the case f o r s u f f i c i e n t l y smooth OA(~) (cf. e.g. MICHLIN (1965)[61]!). Examples: (1.20) i t s symbol is (1.21) OH(~) =  i  s i g n I. The " H i l b e r t transformation" m(y)dy (Hm)(x) : = _1 ~S y _ x f o r
IT =co
(cf. e. g. TITCHMARSH [101 ], p. 120, ( 5 . 1 8 ) ) . 2. The "Calder6nZygmundMichlinoperator (1.22) (M~)(x) : = ( a l + A f / . ) ( x ) ae~ , f(O) eLq(zn) (CMO)" given by
f(~I)
where
(1.23)
~M(~)
188
A. (!) (ii)
G of i n t e g r a t i o n instead of
:
~n , e
g.
: = {xE~n
: Xn ~ O}
kELI(~ n)
WienerHopf i n t e g r a l equations (WHIEs), Fc~ n cones defined by smooth (n2)dimensional manifolds on domains c~ n , like sn and other smooth s e m i  i n f i n i t e (iii)quadrants 0 g~,
~2 : = {xE~2 ' Xl ~X 2 > O} or wedges W(~)::{xE~ 3 , x l + i x 2 = r ~ , ++ = X3ER} or cones w i t h edges: polyhedral domains.
A l l of these are special cases of Definition a given by (1.24) Remark 1.3 : l < p < = , and B. (Tp(A)~)(x) : = (PAlm(p)~)(x) cases" mentioned above we got for Gc]Rn and ~_= LP(~Rn) A E~((3E) , 1.3 : "General WienerHopf operators (WH0s)" : Let P = P2E~(3E) a linear, AE~(]E) , 3 is
Bspace, and
continuous p r o j e c t o r on 3C . Then i t
i n t e g r a l equations" (1.25) or (A~)(x) : = a ( x ) ~ ( x )  ~ k ( x , x  y ) ~ ( y ) d y = f ( x ) E L P ( ~ n ) IRn  even more 1.4 : Equations w i t h "generalized t r a n s l a t i o n (A~)(x) : = ( g i l x ~ A ( X , ~ ) . F y ~ ~ + ~ ) ( x OA(X,~) = j =s a j ( x ) .OA.(~ ) , aj l O being m u l t i p l i e r . A and B lead to important classes of s i n g u l a r i n t e g r a l "classical Cauchytype s i n g u l a r i n t e g r a l ~(~ , e. g. symbols, and V being continuous c o e f f i c i e n t s being a completely continuous on ~n i n v a r i a n t operators ( G T I ~ ) "
~A
operator on ~
Combinations of (j)
equations on manifolds
30 = ~ c ~ , a Ljapounovcurve: equations"
(1.27)
(K~)(x) : = a ( x ) ~ ( x ) + ~ i f k ( x , y ) ~ ( y ) d y x  y
= f(x) with
189 (jj) f(x,e) C. CMOsor singular integral equations with "Giraud kernels" with instead of ac{ or f(o) eLq(En) in a(x) and
Wm'P(IR)
or
Wm'P(IR+)
(A D )(x) + (Vm)(x) = f ( x ) e ~ = ~#
K are onedimensional generalized Llconvolutions (cf. eq. (1.25)), n the A~ = FI OA(X,~). F , ue]li , are a r b i t r a r y generalized t r a n s l a t i o n i n v a r i a n t operators and V : Wm'P(G) LP(G) a compact operator, respectively. A l l these operators, of course, are special versions of "pseudodifferential operators", a term introduced by KOHN & NIRENBERG in 1965 [ 5 2 ] in standard use as (1.30) L : 3 I ~ defined by for mE~ ~ into a series mk. Here , and, since then,
(Lm)(x) : = (F 10L(X,~)Fm)(x )
where
OL(X,~ ) may be expanded asymptotically with respect to ~ _k(X,~) = t ~k ~ homogeneous with respect to k(X,~)
co
of functions o _k ( x , t ~ ) (i.31)
C of degree
for
We shall desist from entering into the discussion of the r e s u l t s on pseudod i f f e r e n t i a l equations since there are many excellent survey a r t i c l e s (cf. e. g. FRIEDRICHS [36], SEELEY [83], CORDES [12], ESKIN [31], KUMANOGO[57]). Here we shall keep close to the lines of SIMONENKO's theory started in (1964, '65)[91,92 ] and shall report mainly on results of our research groups.
Acknowledgement. This paper has been prepared partly during the time when the University of
Regina, Saskatchewan. He wants to thank the Head of the Department, Prof. E. Lo Koh, for arranging ideal wo~ki~ conditions and for the great hospitality there. The author wishes also to thank Dr. F.O. Speck, TH Darmstadt, for his valuable criticism and substantial remarks during the preparation of the manuscript.
190
2.,~,,,Integral and i n t e g r g  d i f f e r e n t i a l equat!ons of the WienerHopf type
WIENER & HOPF studied in 1931 [ 107 ! certain types of homogeneous convolutional integral equations on the h a l f  l i n e R+ in connection with problems of radiative transfer. They developed the functiontheoretic method, a f t e r applying the Fourier transformation. This is now called the "WienerHopftechnique". KREYN in 1958154] completed the classical theory of WHIEs in LPspaces (2.1) where kELI(~) (W~(x) : = ~(x)  I _ ~
0
k(xy)~(y)dy = f ( x )
E ~+
and
L~(~+) such as Co(~+) for instance, are given and me~.+ is sought. GOHBERG & KRE~N [43 ] extended these functionalanalytic investigations to systems of WHIEs. Making use of the convolution theorem of the Ftransformation arrives at the image equation to eq. (2.1) (2.2) where (2.3) and
A
(1%p~2) one
[i~(~) ] ~+(~)
~(~) : = (FP+m)(C) : : ( F ~ m ) ( ~ )
(2.4)
are onesided Ftransforms which may be continued holomorphically into the upper, H+, and lower, H, complex halfplane, respectively. Here we put
(2.5)
h(x) : :
~f o
k(xy)m(y)dy
for
x < 0
for
x > 0 .
Equation (2.2) a c t u a l l y denotes a "Riemann boundary value problem for the l i n e " which is a special form of
(2.6)
+(t)
= G(t)(t)
+ g(t)
tEF ,
F being a smooth contour c C,G(t), g(t) given data and (t) the unknown boundary values of a (sectionally) holomorphic function ~(z) vanishing for z~. In solving such problems (cf. e. g. the books by MUSHKHELISHVILI (1953) [62 ] or GAKHOV (1966) [37]) one of the crucial steps is the p o s s i b i l i t y of " f a c t o r i z a t i o n "
191
of
^ 1 G(t)  or in our example of [ 1  k ( ~ ) ]  into G(t) = a  ( t )  f t  P 1 < A+(t) ~tp_ z Am(z) holomorphic in # 0 on D , the i n t e r i o r and e x t e r i o r domains of F F "
G along r"
G(t) # 0
one has to know a b i t more than Wieneralgebra, f a c t o r i z a t i o n i.e. Ow(~) : = I  ~ ( ~ ) # 0 Let kELI(~), with on
GELS(?). In case of
the onedimension~
convolution operator,
~ E . The r e s u l t by KRE~N[34] is given in the f ~ l w i n g above) be given. Then the WienerHopfoperator is > 0 . E ~(3C~+) (FredholmNoether) i f f
fEX~as
P+(Ik~)P+ inf
P+ : = x~+Ii~(~)I
IOw(~)I = i n f
I f t h i s c o n d i t i o n holds then (i) (ii) m(W) = max(O,<) , 8(W) = max(O,<) for < > 0 there e x i s t s a base and mok(X~., , k = 1. . . . . <  i , j = 1 . . . . . <1 and ind W = u(W) = K of ker W such t h a t
{mol . . . . . mo<}
~ok E L l ( ~ + ) n C o ( ~ + ) d mo,k+l(X) = (2.10) mo,j(+O) %,<(+0) (iii) (2.11) where (2.12) for < < 0 = = 0
c o n d i t i o n s are
/ f(x).~ok(X)dx = 0 o
k = 1. . . . .
I<l
7 k(yX)~ok(Y)dY
0
= 0
(iv)
192
oo
(2.14)
y(x,y) : Y l ( x  y ) +  { 2 ( Y  X ) + ~ l ( x  t ) Y 2 ( Y  t ) d t
0
Ll(IR)functions
: (I+(FP+YI)(~))(~)
i
lk(~)
.(I+(FP+~2)(~))
The reasoning which leads to theorem 2.1 has been carried over to
the case of ~ + , n m2; instead of ~+ by GOLDENSTEIN & GOHBERG in 1960 [48] where the f a c t o r i z a t i o n applies to the nth Fvariable Cn in ~. In this case A i [ a r g ( l _ k ( ~ l , " " " ,~n))] is always 0 f o r an e l l i p t i c WHO since due to K : : T~ ~n=~ the RiemannLebesgue l e n a , continuously on ~(~i . . . . . ~n)ECo ( ~ n ) and tends to implies that 0 for K depends ~' : = (~I . . . . . ~nI ) l~'I ~
2.2 : Concerning the group B of generalizations {AHBAGJAN in 1968 [7] ] and RAKOVSH~IK in 1963 [ 75 ] t r e a t e d "variable kernel WHOs" (2.16) (Wm)(x) = m(x)  f k(x,xy)m(y)dy = f ( x )
co
E3E+
IR+
n
ai(x ) and
ki(t )
X#~n
sup l a . ( x ) I
1
ielN
kieL1(IR ) such t h a t
A
> 0
in ~AHBAGJAN's paper, while RAKOVSH~IK[75 ] assumes (2.20) m(x)~+ I k(x,xy)m(y)dy LP(N) LP(a,b)
to be a bounded operator on X = LP(IR), ISp<~ , be compact from f o r any f i n i t e (2.21) (2.22) [ a,b] c]R , lira
X~_+~
Ik(x,t) and
f o r large
+x > 0
n+(x) = 0
(2.23)
inf
~elR
193 Here
A
__
oo
(W~)(x) = ~(x)

i~lai(~).~nki(xy)m(y)dy
+
= (lW=lm(x) + (wlm)(X)
= f ( x ) E3E~+
where IW~ c~(3E+) and even boundedly i n v e r t i b l e f o r ~ = 0 ( i . e . always' for n~2!) 1 and W is the sum of a compact operator and one of "small norm", so being E~(~+) too. Thus the index of W is the same l i k e that of the f i r s t term.
2.3 : WHOs, p a r t i c u l a r l y for ~+ , have been studied also f o r the whole scale of SobolevSlobodezkispaces ws'P(~+) and w~'P(~+), the functions of ws'P(R) : = { f ~ ' : FI(1+I~I2) s/2 FfcLP(~)} ; l~p<~ , sE~, also called "spaces of Bessel p o t e n t i a l s " , and the subspace of them having supports, supp f , in ~+ while W ' P ( ~ + ) denotes the r e s t r i c t i o n s P+f of fEwS'P(~) to ~+ s Due to the Sobolev embedding theorem i t is wellknown (cf. e.g. TALENTI (1973)[100], p. 28) that Wo' s,2,~+)^wS,2(o ~  ) = {0} for s ~  1 / 2 but span { ~ , 6 ' , . . . , ~ ( n ) } f o r ,s,2,~ .s,2, s < 1/2 and the largest integer n <  s  i / 2 while wo ~ + ) m w o (~_) = wS'2(~) for Is~ < 1/2 and = { mEwS'2(~) : m(k)(o) = O, k = 0 . . . . . n  I } for n1/2 < s < n + 1/2, nE~. The defect pair WHO then depends on the number n d e t a i l s the paper by TALENTI, pgs. 63  71!) 2.4 : first The theory of WHOs on the h a l f  l i n e has been generalized to operators of the kind where the symbol is given by C() vanishing at i n f i n i t y and to more (~(W),B(W)) of the onedimensional being related to nI/2 < s < n+1/2 (see f o r the
general " n o n  e l l i p t i c or nonnormal WHOs" whose symbols may be written in the form (2.26) with ~jER ~i < N ~~. n. aw(C) = ( c  i ) n~ G'(C)(~~~) G+(C)'j= 1 (~TT ~) J ~ being d i s t i n c t , n , NE]~ ; n j E ~ . Considering the d e t a i l s to these 1965 by
SAMKO [78] and mainly by PR~SSDORF (1965, '67, '69) [68~69~70 ] look at PR~SSDORF's book (1974)[71]! TALENTI (loc. c i t ) is involved mainly in WHOs of the f i r s t kind (pgs. 71  77) which play a dominant role in the study of mixed boundary value problems in ~+2 (cf. e.g. PEETRE (1963)[64] , SHAMIR (1962, '63) [84,85 ] and
194
f o r the higher dimensional case: ESKIN's book (1973)[31]!) 2.5 : In quite a s i m i l a r way "twopart composite convolutional equations" and t h e i r
adjoints the "dual i n t e g r a l equations" may be treated via the Ftransformation and the Riemann boundary value problem:
0
(2.27) where
u(x) = u I
for
x < 0
and u(x) = u2 f o r
x > O, r e s p e c t i v e l y .
(2.28)
(.~)(x)={
A good survey on t h i s subjects with a p p l i c a t i o n s to other dual i n t e g r a l equations is provided by the book by ZABREYKO et a l . (1975)[108] in Chap. V I I I , 4  6. eqs. of the WienerWe are now going to review the r e s u l t s on i n t e g r o  d i f f e r e n t i a l
Hopf type, i . e. some kinds of general WHOs acting between suitable Sobolev spaces, viz.
m
(2.29)
(Lm)(x) : =
x > O.
BANCURI in 1969 [3] treated only the case of constant c o e f f i c i e n t s a , b , c assuming b c 1 or  O , k ELI(]IR), ' f E L I ( ~ ) given and q)~'O m ) ~L =~+ ~ sought, ~,l( i.e. m. . . . . m(m)ELI(]R+) and m(+O) . . . , ~(mI )~ +0 ) = O. He applied the F  t r a n s f o r mation and a r r i v e d at the Riemann boundary value problem (BVP) a f t e r some manipulations: (2.30) Set (2.31) ~+(~) : = am.(i+~)m~+(~) ^_ ^
(~1 : = h  ( O
A and
(2.32)
G(~)
(2.33) (2.34)
g(~)
^
: = f ( ~ ) , g(~)
m oo
where f k (xy)D~m(y)dy)(~)
h(~) : = (Fx]R ( x ) . z
!J=O 0
(2.35)
~ () = G(~)~(~) + g(~)
in ~]~(]R)
195 which is e l l i p t i c i f f G(~) # 0 on JR, GERLACH (1969) [38] admits a l l the Kre~'n spaces and the subspaces )[~m) : = {q)e)E+ : D m ~ f o r u = 0 . . . . . m} or ]E[(+m) : = {mE~(+m): (Djm)(+O)= O; u j = 0. . . . . ml}. by w r i t i n g (2.36) F i r s t he t r e a t s the constant c o e f f i c i e n t s ' + il
m
D = (Dil)
(Lm)(x) = P+
~ (a , l + k p , l ~ ) ( D  i l ) U m ( x ) p=o = (Gmm)(x)
~(x) : = ( D  i l )  m m ( x ) he gets
(kgm~)(x) = (W+Vm)~(x)
= f(x)c3E+
is a L l ( ] R )  k e r n e l
WE~'(~c+) ~
~w(~) # 0 on
Ow(~) = OA(~)'gm(~ ) =
Then Gerlach treats the case of continuous c o e f f i c i e n t s ap(x),b ( x ) , c p ( x ) using a simpler version of RAKOVSHCIK's compactness theorem (1963)[75] of convolutions multiplied by functions: Put m ,Bp,yu f o r the l i m i t i n g values f o r ap,bp,c as x + ~ and rearrange eq. (2.29) to read m (2.40) (km)(x) = P+ ~ (~ l+~p~ .k~*)(D~m)(x) p=o
m
( : : ~)m(x)) +
+ (b (x)6p)k * c ( y ) ' I
l](DUo)(x) .
is a WHIDO of the type above and the last term of eq. (2.40) being 3(~(m) into ~F.. The same argument applies to the subspaces ~ ) o W 'P(~+) m for l<p<~.
compact from
196 GERLACH's r e s u l t s may be summarized in Theorem 2.2 : on ~ + , ~ eq. (2.29) (2.41) Let a ,b , c C ( ~ ) , ~ k f~Ll(~)_ for u= 0,1 . . . . . m with am(X ) ~ 0 in
: = lim a~(x) ; 6 similarly___defined. Then the WHIDO L is x + + ~ e ~ + , ) + ) ~ orY ~ ( ~ ) , ~ o + ) E ' i f f i t s "main symbol" ~ ) : = m ^ z [a +8 ~ "k (~)]~ ~ # 0
~=0 ~ P p
on N~
I f t h i s is true one has the r e l a t i o n s (2.42 a) (2.42 b) and (2.43 a) (2.43 b) ind L = v(L) = v(~ = < + m/2 = K  m/2 0 s e(Lo) ~ ~(L) s ~(Lo) + m 0 ~ 6(L) g 6(Lo) where Lo : = Ll~(m )
~0+
ind Lo = v(Lo) = v ( ~
with the windingnumber (2.44) <(L) = ~(Lo) = ~ L a r g L is given by :(~(Azl),6(Azl)) ~+ to # (0,0)}. R n+. I f Ow(~)]~=_~
s(L) : = {zE : z = q L ( ~ ) , ~ e ~ } u { z c C
one takes (nl)dimensional Ftransformation with respect to x' : = (x I . . . . . Xn_ I ) one a r r i v e s at an equation in the remaining v a r i a b l e x n > 0 but containing now ~' : = (1 . . . . . ni ) equations (2.46) where to ~+. as parameters. So i t is quite natural to i n v e s t i g a t e
P+A(e~Dn)m+(Xn) = f(Xn) , x n > 0 8' : = ~ ' / I ~ ' I Here A(8',Dn) has been f i x e d and P+ denotes the r e s t r i c t i o n operator
is defined as a p s e u d o d i f f e r e n t i a l _i ^ : = (F n A(e',~n)Fnm(Yn))(Xn)
operator (PDO) by
(2.47)
A
A(e',Dn)~+(Xn) A(e',(n)
where
is homogeneous of degree
m with respect to
(n
general one ( c f . the work by VlSHIK & ESKIN (1965,'67,'73) ~i03,104,105] RABINOVI~ (1969 , ' 7 1 , ' 7 2 ) [ 7 2 , 7 3 , 7 4 ] (1971,'73) [17,1~ and others!). , BOUTET DE MONVEL ( 1 9 6 9 , ' 7 1 ) [ 5 , 6 ] ,
197
2.7 :
WHIEs and WHIDEs may be considered not only for Cvalued or sNvalued
functions but in a more general context as Bspacevalued equations. Then the n problem for ~ + , n~ 2, may be f i t t e d into the theory too. Concerning a general theory of operator WHeqs. FELDMAN investigated several cases (1971)[32,33,34] in connection with problems of radiative energy transfer. GRABMOLLER(1976, 1977) [49,50] discussed such i n t e g r o  d i f f e r e n t i a l equations on ~+ of f i r s t order with a linear, closed operator (2.48)  A generating an analytic semigroup involved = f(t)E~.~
where h o , h l E L l ( ~ )
a r e f l e x i v e Bspace and
denotes the strong d e r i v a t i v e , the integrals to be understood in the Bochner sense of LP(R t;31[+), %aEt. He is mainly interested in the asymptotic behavior of the solution as t ~. 3. Compound integral and i n t e g r o  d i f f e r e n t i a l and Llkernel type on ~n equations of the princ!pal value
MICHLIN [ 60 ]introduced in 1948 the notion of the symbol f o r singular Cauchytype integrals along curves r c and also f o r operators on ~ n , nm2 . He and mainly CALDERON & ZYGMUNDstudied the mapping behavior of the CMOs since 1956 (cf. e.g. [ 8 ]). The f i r s t systematic treatment of the corresponding integral equations probably was published in MICHLIN's book (1962) whose English translation appeared in 1965 [61]. A more recent account, also on i n t e g r o  d i f f e r e ~ t i a l equations with CMOs as c o e f f i c i e n t s , may be found in Chap. IX of the book by ZABREYKO et al. (loc. c i t . ) . AGRANOVI~ (1965) treated equations of the following type in his extensive survey a r t i c l e [1 ]: (3.1) (Am)(x) : = s (MuDUm)(x) + ( T ~ ( x ) : f ( x ) l~l~m ~n replaced by
n ~+
as an operator
A : wm+~'2(~ n) W~'2(~ n ) , or
or a smooth
compact manifold ~O . The symbols (3.2) oM (x,~) : = au(x) + (p.v. Fy~ fu(X'ly_]_) )(~) y lyl n
> nI are assumed to be EcP(]R n ,Hq(sn) ) where pc]No and q ~ suchthat, by Sobolev's embedding theorem, they form an algebra of continuous functions on ]Rn>~ En which are homogeneous of degree zero in ~ . He shows that to every such function o(x,~) there corresponds a characteristic f(x,O) EcP(IR n,Hq'n/2(sn )) (theorem 7.12). The operator T is one of order almost mi which would be n compact for a compact manifold ~9 instead of ]Rn or JR+ by Rellich 'S c r i t e r i o n .
198
AGRANOVI~ proves a couple of theorems which give necessary and s u f f i c i e n t conditions f o r A to be FredholmNoether by means of the e l l i p t i c i t y condition of the symbol or the existence of a  p r i o r i estimates (theorem 12.1). He obtains then the wellknown properties f o r e l l i p t i c of operators, such as r e g u l a r i t y , s t a b i l i t y with respect to parameters etc. (cf. his theorems 12.2, 12.3, 12.4L). But, in the case IRn or IR+ he does not give regularizers in the sense of theorem i . I above. n SEELEY[82] investigated at the same time (1965) singular i n t e g r o  d i f f e r e n t i a l We are not going to enter into t h i s detailed material but j u s t want to give two d i f f e r e n t approaches: one relying on DONIG's work in (1973,'76) [19, 21 ] and the other on SIMONENKO's (1964,'65)!91,92] , RABINOVI~'s (1969'72)[72,73,7'4] and SPECK's approach (1974'77)[96,98]. (n=i) (3.3) In 1973 DONIG[ 19] treated the case of ~R with the singular i n t e g r o  d i f f e r e n t i a l operator (SIDO) m (A~)(x) : = ~ {au(x)l+b (x)H+c (x).ku~}(DP~)(x) = f ( x ) p=o ws'P(IR) , sEIR, l<p<~, with c o e f f i c i e n t s is
on SobolevSlobodezki spaces
am(X), bm(X), Cm(X)CC(]R ) and a (x)a (=) etc. CLo(]R ) = {~L=(]R): mes {Ix~i>n : I~(x)l > E} +O,n + ~ , for a l l c > O} and where fEWsm'P(~R) A through
A
(3.4)
:=~u{~}u{+=}. ~A(X,~) # 0 on ]RxIR , where Using the Bessel potential operators jm : = F1(1+I~I2)m/2F he gets the relations
iff
(3.5)
Dmjm : ( i l l ) m (l+rmw)
(A dm~)(x) = (B'm~)(x) +
(3.7)
B"
: = (a
I + b H + k ~ ) " { (iH)m u ~ I
for for
~ = m O~ u<mI
and some r~ELI(1R). He succeeded in c o n s t r u c t i n g a bounded i n v e r s e operator CmC~(LP(IRn)), l<p<~, in the form of (3.8) C = Mml(iH) m + hl,m~ + h2,m~H m
199
where M : = am.l + bm.H, H is the Hilbert transform and hl, m , h2, m E L I ( ~ ) m e x i s t due to the WienerLevi theorem. He gives an e x p l i c i t formula, though
a ^
hl,m(~ )
and
C then implies m
(3.9)
with a
c o e f f i c i e n t operator with ~m : ~m(=)' Bm : = bm(=) is s p l i t off and is inverted with C as the inverse (A .J m) m on L P ( ~ ) . The product with the perturbed terms, and p a r t i c u l a r l y the lower order ones, gives a compact operator on LP(L~) (Rakovsh~ik's r e s u l t from 1963!). This leads to the coerciveness inequality:
(3.10)
with a certain
LP(IR)) having an ind A = v(am(x)+bm(X)H ) which may be calculated from the a_iblx~l~j~ . In case the data are smoother, e. g. winding number K : = [arg ~
oo
am, bm~S(~R)~sE~
c
, av,bvccsl(~ )
and a (s)
b(S)CLo(]R)
fEwS'P(]R) smooth solutions mEWm+s'P(IR). After discussing the dual operator on W m,p (IR), 1/p' +l/p = 1,
m v
(3.11)
(A*~)(X) : :
S (1)~DV{au(x).l  H b ( . ) . l + k v ~ } ~ ( x )
p=O
where ki(x ) : = kv(x ) i t is possible to show that A and A* may be extended continuously  for s u f f i c i e n t l y smooth coefficients a (x), bv(x )  from Wm'P(~) to J)Lp(N) and from wm'P'(N) to j ) ~ p , ( ~ ) . DONIG in 1976 [ 21]extended his reasoning in order to include the ndimensional version of the singular i n t e g r o  d i f f e r e n t i a l eq. (3,3) to the case of coefficients M + K where the M are CMOs and the K of L l ( ~ n ) . kernel type, both being xdependent. In contrast to AGRANOVI~ in (1965)[ 1 ] h e succeeded in arriving at an e x p l i c i t r i g h t and l e f t regularizer in the case of e l l i p t i c operators A . I t is impossible to give all details here, but his main result is formulated in the following Theorem 3.1 : Let n ~2, l<p,q<= I ni q go (q) : = > ~o(q). such that for i<o<_2 .q ~ pP1
n_@_
for
q~2
and
200 Furthermore assume that S f~(x,~)d~ = 0 for a l l a (x)EC(~ n) , f (x,o)EC(~ n, Wg'q(sn)) xEIRn and k (x,t)EC(IRn, L I ( N n ) ) , with
I~I < m.
f (x,~)
(3.12) (Am)(x) : = lu!~ {a~(x)(D~m)(x) + p.v, mn~ u Ix Ix[yl(Dum)(y)dyyl'' + ~m~ + f k~(x,xy)(D~m)(y)dy} IRn be defined by (3.13) = f(x)ELP(~R n)
f (x,  ~
aA(X,~) : = I~ IS {a (x)+ p.v.F Y~+~( ~m symbol" by f~(x,T~ F) ]y[n ) ( ~ ) } ~ to be a iYl n (~) + (FY'+~k~(x 'Y)) ( ~ ) } ~
Then putting
Fourier m u l t i p l i e r I. A is e l l i p t i c
(3.15)
(3.16) 2.
or
A is coercive on Wm'P(~R ) , i . e . n T > 0 and a compact seminorm l.] such that eEWm'P(IR n ) ,
L~I
for a l l
(3.18) where
ind A = ind (
M u) = 0 and the R
R = R1 ~
1.
...'R n
~n
are powers of
201 I f one of the conditions holds then a twosided r e g u l a r i z e r by (3.20) where the symbol (3.21) and (3.22) ~ B : = F i (OA(~).(l+I~ 12)m/2)iF DONIG (1977)[ 22] applied his method, in case of ~(z) , z c C  i ~ , i~. n = 1 , to give transmission B = B~.M M M M =, are CMOs. M is a bounded r e g u l a r i z e r to 0 , lui=m M "Ru with B may be calculated
aAo(x,~)l~Im M : = ~
f (@) : =
Remarks : 3.1 :
conditions f o r the Fredholm property of a transmission problem f o r s e c t i o n a l l y holomorphic functions which have i n t e g r o  d i f f e r e n t i a l On the other hand (1974) [20 ] conditions on the common boundary he solved the Cauchyproblem (3.23 a) (3.23 b) with e l l i p t i c ~ + A(t)~ m(x,+o) operators t = f(x,t)EC([o,~],L2(N n))nCz((o,~],k2(N n)) = mo(X)cL2(Rn) A as before of the compound Llkernel and CMOtype, now
even depending on
parabolic evolution equations. The p r i n c i p a l symbol is assumed to be "uniformly strongly e l l i p t i c " : f (x,t, (3.24) for all Re(l) m x c~ n z {a ( x , t ) + (p lul =2m ~ .V.Fy~ lyln 41 ) ) (~)}~ > cl~ 12m =
and tE [o,~].
The r e s u l t s are quite s i m i l a r to those in FRIEDMAN's book (1969)[ 35 ] 4. Simonenko's theory fgs. generalized t r a n s l a t i o n i n v a r i a n t operators
We want to give now a short outline of SIMONENKOs approach which he established in 1964,'65 [91,92] (1974'77)~96,98] iff ABE~T(~L,I~) and which was applied by RABINOVI~ (1969'72)~72,73,74] and SPECK in the theory of GTIOs. For abbreviation we shall w r i t e for A,BG~(3(,~) where ~,~ for simplicity) A~B : are Banach spaces. ATKINSONs
Ac~(3E) i f f there are RL, RrE~(~ ) such that R~A~ARr~I. In what follows we shall mainly be interested in the spaces ~ = Wm'P(IRn), mcIN , l<p<~ , and t h e i r o
202
1 1 W m'p ' (JRn ) , ~ + ~, : I. So, for the following d e f i n i t i o n s and results, l e t
duals
us assume ~c~_c~' being dense in the topology of the bigger space. Let the translation operator ThE~(~ ) and the multiplication operator m(x).l with a C~(]R)  f u n c t i o n ~ be a bounded operator on ]E , too. ]Rn may stand for ~n or ]Rn (see sec. 0). Definition 4.1 : (i) iff m1.A~2.1~O AE~(3E) is said to be "of local type (with respect to with ~im2 = 0 . We then w r i t e ]Rn) ''
AE.A,(L~]. xo ( i i ) A,BE~(~E) are said to be " l o c a l l y equivalent at Xo~]R , written as__ A~B , n i f f for all ~ > o there exists a neighborhood %(Xo) and an ~IzEC=(]Rn) with such that
for all
ml,~2~C~(]R n)
mlL(x) m 1 on ~ (Xo)
(4.z)
(4.2)
inf II V~0
~I(AB)VlI~(z)
< ~
and , too.
if'', written as A E ~xO(~) ( i i i ) AC(~) is said to be " l o c a l l y Fredholm at XoeR i f f there exists a neighborhood ~(xo), and a pair of operators RL,x o and Rr,xo such that (4.3) (R~,xoAl)w~.l~O and ~ .l(ARr,xol) ~ 0 .
Remark 4.1 : I t may be shown (cf. e. g. SPECK (1974)[96], p. 50) that in case of ~[= LP(A n)___pproperty ( i ) is equivalent to ( i ' ) [ A , w . l ] : = A ~ . I  ~.I A ~ 0 for all ~CC~(R n) and ( i " ) X E . A xG.I~O for all E , G c ~ n with EnG = ~. For this case also ~ # C ~ ( ~ n) may be replaced by ~ in ( i i ) , ( i i i ) , see also RABINOVI[ (1969)[Y2]. SIMONENKO in 1964 [91]proved the following theorems. Theorem 4.1 : Theorem 4.2 : Let Let AEy~(X). Then A E ~ ( ~ ) A,B c~(~) iff AE~o(~_) for all . Then XoEIRn.
Xo~
A~x (3~)
iff
Bc ~Xo(~"
Some of SIMONENK0's additional results had been generalized by SPECK (1974)[96] ,viz. Theorem 4.3 : AE~Xo vertible. Theorem 4.4 : Let X = LP(~ n) , l<p<~ , and Let A E~0E), where ~_= Lp ( ~ n ) , xo E ~ I ~  R n lz:Jp<~ or iff A .3E= Co(~n ). Then is (continuously) in
203
(4.4)
Ixy
a compound CM and Llconvolution operator, (cf. eq. (1.22)); then finite point xo E ~ n iff al + Af/. is i n v e r t i b l e .
Remark 4.2 :
a) Z :
O
[#~{n),
These results c~rry over to more general situations as e. g~ ~n 1<p<~ , A = r ~ ( ~ ) . P , c being ~ stabilized, i . e. c = o
%,
being homogeneous of degree zero and co tending to zero at i n f i n i t y , or b) 3C_= L2(R n) , ~ being an a r b i t r a r y L'function (cf. SPECK [97]!)~ Thus we are led to the following generalization of Definition 1.4. D e f i n i t i o n 4.2 : ( i ) A~(3E) is called a "generalized translation invariant
operator (with respect to ~T~),, i f f for any XoE~ there is an AxoC~(~) NA (3~) ~ch that AX~A Xo I then(ii) I f A is a GTIO "enveloping" the family {Ax}xE~n where Ax ~ = F CAx(~)F,
~n ~
is called a "presymbol of
A ,,
is a homomorphism of the space of GTIOs onto the quotient space ~/~'o with kernel O~, the compact operators (the sequence {o} + d~ GTIOs +~/I~"o ~ {o} is "exact"). I T Thus d e f i n i t i o n ( i i ) makes sense. The main results in this context may be summarized in the Theorem 4.5 :
Xo
(SIMONENKO (1964)  SPECK (197477)): N A(~) for all XoC~ "~. Then the following conditions are
for all
XocNn;
p = 2 at the same time holds sup ess inf l~(x,~)  Co(X,~)i > O. ~oe ~o (x '~)e~Trxmn
204 (v) ess i n f l{(x,~)[ > 0 ( x , ~ ) e ( ~  ] R n ) x IRn has to be continuous from ~'T~_ ~Rn into L=(~n).
~l(xo,~ )
( t h i s leads to more complicated conditions). Applying t h i s to the compound poledependent Calder~nZygmundMichlin and L1integral operators on ~ = L2(]Rn) given by xy (4.5) (Am)(x) = a(x)m(x) + p.v. I in ]Rn IxY
f(x,
m(y)dy
+ f k(x ,xy)m(y)dy
]Rn
one gets the Fredholm c r i t e r i o n by the symbol conditions (4.6) i n f laA(~,~)i ~ 0 ~e~ n min xE~ n inf I o ~EZn a ( x ) I + A f ( x , . ) / .
(4.7)
(~)I > 0
These are exactly the same conditions l i k e those got by DONIG [21] in the case m = O. As we had seen he proved furthermore the equivalence to coercive inequalp = 2. Due to the kind of the characterization theorem for xo there occur restclasses of functions in the d e f i n i i t i e s , while SPECK [97] gave the complete extension f o r a r b i t r a r y (nonstabilized) GTIOs in the case of Ac~"x tion of the symbol. Now, a l l the considerations above may immediately be generalized to integrod i f f e r e n t i a l equations (4.8) (Am)(x) = l ~z~ l (ApDUm)(x) = f ( x ) E at f i n i t e points
where the c o e f f i c i e n t s A are generalized translation i n v a r i a n t operators according to d e f i n i t i o n 4.2 ( i ) and ~ c ~ m ) c ~' be d i s t r i b u t i o n spaces such that D~mc~ f o r ~E~ o , 0 ~ lul ~ m. F i r s t we notice that the concept of t h i s section A : ~ ~ operating between d i f f e r e n t can be completely transferred to the case of
Bspaces. Then choosing l~ = ~C(m) we can see that the Bessel potential operator jm = ~1(1+i~12)m/2.F wi~l reduce eq. (4.8) to one in ~ . The essential f a c t here is that the js are not only t r a n s l a t i o n i n v a r i a n t operators from ~ ( t ) = wt,p(~n) onto ~ ( t + s ) = wt+s,p(~n) for t,sEIR, l<p<~ but also of local type with respect to ~ n , cf. RABINOVI~ (1972)[74 ]and SPECK (1974)[96]. So we can state the
Let
A is a generalized translation invariant operator form ws'P(]R n) to wt'P(]R n) i f f ~ : = j  t A j s is one on LP(~ n) (with respect to ]Rn!) AE~(wS'P(]Rn), wt'P(IRn)) i f f ANc~(LP(]Rn)),
I t turns out that "generalized convolutional i n t e g r o  d i f f e r e n t i a l operators" like in eq. (4.8) are generalized translation invariant operators from Wm'P(]Rn) into LP(]Rn), l<p<=, with respect to ] ~ n So i t is easy to reformulate the Fredholm conditions for such operators. Theorem 4.7 : Let I t is where (cf. RABINOVI~ [ 7 2 ] , p. 87, Th. 4.1; SPECK [96] , p. 89) u A : = i ~ l ~ A D be a generalized convolutional i n t e g r o  d i f f e r e n t i a l c~:'(Wm,p(~n), L P ( ~ n ) ) , {Ax } . xem n l<p~, iff Ax j m E ~ x ( L P ( ~ n ) l for all is the family of operators
operator. xE~ n
(operator with "frozen c o e f f i c i e n t s " ) . I t is necessary that A= be b i j e c t i v e and i t is s u f f i c i e n t that for all (4.8) xE~ n. This holds for >0
in# I #~rn~A(x,~)'~U(l+lI2)m/21 c~ n ~
~cR n
i f p = 2 in which case the pseudoinverse of AJm can be constructed as an enveloping operator to the local inverses (AxJm)IE#NA(L2(~n)). For p # 2 the local existence of these operators must be assured (cf. theorem 4.6). Remark 4.3 : The concept of local equivalence of operators has been generalized by SIMONENKO (1964,'65)E91,9~ from the very beginning to the following one D e f i n i t i o n 4.3 : Let , Y be Hausdorffspaces being homeomorphic by ~ : X Y and l e t (T~u)(x) : = (u,~)(x) be a nondistorting transformation for all ueLP(Y) , l<p<~ , fixed. I f Yo = ~(Xo) and ~L(Xo)C and ~(Yo)CY are homeomorphic neighborhoods then AeA(LP(x;~)) and B~A(LP(Y;~)) are called " l o c a l l y quasiequivalent with respect to the pair Xo' Yo"' written as AXe9 Y~B i f f
(4.9)
T _IP A
Yo P~B P~r
P~ = ~.I and P~= 7~I,
in the sense of d e f i n i t i o n 4.1 ( i i ) with space projectors respectively, u and n denote complete respectively.
206 D e f i n i t i o n 4.4 : Let ~ = LP(]Rn), l_~p<~ , and NEj, j = 1. . . . . N, be a f i n i t e set of d i s j o i n t measurable sets in ~R n such that u E. = ~ n I f A j , j = 1. . . . N , are generalized t r a n s l a t i o n i n v a r i a n t operatorsJo I J ~ (with respect to ~I~) then
N
(4.10) is called
W : = N
a "component GTIO" or an "Npart composite WHO". I f Aj, j ~WN(X,~) = 1. . . . . N, then the "presymbol of : = OA( x , ~ ) j for (x,~)CEj~<IR n. WN"
presymbol of (4.11)
is def~lned by
Theorem 4.8 : (SIMONENKO(1967)[94], p. 1322) : LetW N be a component GTIO as above where Ej = Fj GTIOs of L l  t y p e , and IR~ = ~ n Then W N ~ (4.12) where (4.13) Aj : = [ ( ~ N } ~ j ) ~ W N n]u[Fj~{~}] equals zero. min inf j = l . . . . . N (x,~)cAj IOWN(X,~) ! > 0
, j = l . . . . . N.
Remarks: 4.4 : This theorem can be generalized in various ways  e s p e c i a l l y for p = 2  by the following more general assumptions:l).Let Ajx , the l o c a l l y quasiequivalent t r a n s l a t i o n invariant operators in xcIR n have symbols ~Jx = ~Jx+ ~jx where 'fjx are " r e l a t i v e l y t h i n " L ~  functions instead of being EFLiCCo(]Rn), i. e. (4.14) sup l~oi~2r ~ l~jx(~)Id~ = o ( r n) I~ o Isr for r ~ ,
co
mes(EjFj)n{~cIR n:I~O~l
~ I} 0 for
l~oI
where the Fj are smooth cones, cf. SPECK [98], too, 3). Let r j be piecewise smooth cones, cf. ~!EISTE~ & SPECK [58]. 4).Assume that l o c a l l y at every 9oint xE~ n more than two sets 4 . 5 : ~j are allowed to i n t e r s e c t , cf. HEISTER & SPECK [59],
SIMONENKO (1964)[ 91] generalizes to composite WHOs with space projectors where the Ej are domains c~ n Aj bounded by smooth Ljapounov area and the operators are singular having homogeneous
pEjE~(L2 ( ~ n ) )
manifolds of f i n i t e
207 symbols (or symbol matrices) of ( p o s i t i v e ) order proves that the symbol ~w(X,~) : = ~A(X,~) values as x x oC~Ej for 0 being continuous on ~n ~ He
(x,~) c E j x ~ n
and t h e i r l i m i t i n g
from both sides has to be zero and a certain indexconin order that W be FredhlomNoether. This can be done N xo ~Ej onto
d i t i o n to be f u l f i l l e d
by reduction to a twopart composite problem l o c a l l y at every boundary point since the quasiequivalence is established by the local mapping of ~ = ~nl.
boundary value problems, f o r smoothly bounded domains. SIMONENKO's method has been applied by RABINOVI~ in (1969)[72, 5] to boundary value problems for generalized convolutional i n t e g r o  d i f f e r e n t i a l equations  and systems of such  in semiinfinite domains GC~n with smooth boundaries behaving l i k e a cone f o r large distance from the o r i g i n . Again the nonvanishing of the symbol on AG : = [ # N ~ R ~ ] U [ G ~ { ~ } ] is necessary and s u f f i c i e n t for (4.18) PGA(X,D)PGm + TPGm = fcHS~(G)
to be FredholmNoether. Here PG denotes the r e s t r i c t i o n operator on s+c,2 (G) and TE~(Hs+c (G), Hsz(G) where ~s+~(G) is the closure Hs+C (G) : = W of C~(G) with respect to For the case of Fredholm property, on II.llHS+~(G)norm.
G = R n , or a bounded domain, a thorough discussion of the n p a r t i c u l a r with boundary conditions or potentials carried
V
[17,18]. RABINOVI~ (1972)[74] then studied pseudodifferential operators on classes of noncompact manifolds with boundary conditions. Quite recently CORDES (1977) stQdied C*algebras of e l l i p t i c boundary problems [10]. Shortly before he derived a global regularizer  or even parametrix  to pseudodifferential operators on R n (1976)[14].
208
5. WienerHopf type integral equations with strongly singular kernels equations having smooth symbols on
G~R~admittingc.l~x for the xdependence of the factors or kernels more general domains or having symbols being continuous on IR~ _ { 0 } n but x varying in the whole of ] ~ n We want to look now into equations combining stronly singular kernels, of the Cauchy p r i n c i p l e value or CalderonZygmundM1chlin type, with piecewise constant coefficients on ]Rn. The simplest case arises from the wellknown x " a i r f o i l equation" +I (5,1) _1 f q~(y)dy 1 'y  X = (m(1,1)m)(x) = f ( x ) , XE(1,1)
J
involving the " f i n i t e Hilbert transformation Hr , . j I~"" BETZ in (1920) [,4 ] gave l an inversion formula for s u f f i c i e n t l y smooth f E C # o c ( (  1 , 1 ) ) N L I ( (  I , I ) ) 0<),<1, by (5.2) m(x) =     ~ _ 1
"
with a r b i t r a r y
H2 = I
acting on L P ( ~ ) , l<p<~ , or C~oc(~) following from eq. (1.21). I f one wants to solve eq. (5.1) above by means of the CarlemanVekua functiontheoretic method introducing give for (5.3) and T(H(1,1)m)(x)
(5.4)
#(z)
g'~3 _1
1 +} my_~d z
[1,1]
r = IR:
~+(x  G(x) =
~(x)
for
xE(l,1)
xE(1,1) xcIR[1,1]
~+(x
So eq. (5.1) is equivalent to the Riemann boundary value problem with a piecewise continuous factor only:
(5.5)
XE~{1,1}
+I
209
T
(5.7) g(x) :
1 f(x) 0
for for
xe(1,1) xE~[1,1]
G at
x = i
solution given by eq. (5.2). This f i n i t e Htransformation, and the s e m i  i n f i n i t e as w e l l , have attracted the attention of many mathematicians up to now concerning one and multidimensional integral equations involving them or t h e i r multidimensional counterparts. While the Russian school around MUSHKHELISHVILI (cf. his book (1953)[6211 has applied the LPtheory, in mainly looked into Hoelder space solutions by the classical method TRICOMI in 1951 [102], NICKEL in 1951 [63], and SUHNGEN in 1954 [95] the l a t t e r case by a transformation which diagonalizes the f i n i t e Htransformation. KOPPELMAN & PINCUS in 1959 [53] and J. SCHWARTZ in 1962 [81] derived spectral representations for H(_I,I) on L2((I,I)). WIDOM in 1960 [106] and SHAMIR in 1964 [86] studied singular integral equations and systems of them, respectively, on measurable subsets Ec~ and on ~+ or [  I , + i ] , respectively. They treated the LP(E)  and ws'P(~+)  , s ~ O, l<p<~, cases, respectively, but were apparently not aware of SUHNGEN's work before. CORDES & HERMAN in several papers, (1966) [15], (1969) [11] introduced Banachalgebraic techniques, the Gelfand theory, to singular integral operators in of the l a t t e r paper KREMER in his thesis (1969) [55] treated the equation (5.8) (W,)(x) : = a(x)~(x) + . o y  x + c ( x ) . f k(xy)~(y)dy o = f(x)EL2(N+) c(x) z O, to L2(~+) of purely Cauchytype and of compound ones with additional Llkernels. IndependenCy
with the symbol calculus by GERLACH & KREMERin (1972,'73)[39,40 ] with piecewise
Following quite a different line G H E G & KRUPNIKstarted in (1968) [44] to OBR study Cauchytype singular integral equations along curves rc{ continuous coefficients anew and introduced 2 x 2matrix symbols for scalar equati~ons in 1970 [45]. The whole theory, even under much weaker conditions on the coefficients a ( t ) , b(t), tEr , in the equation (5.9) (K~(t) : = a(t)~(t)+b(t).(Sr~)(F ) = f(t), tErc~ ,
has been displayed in a l l d e t a i l s in GOHBERG & KRUPNIK's book (1973) [47]. DUDU~AVA applied t h e i r theory to WienerHopf type equations and compound equations with symbols having weak smoothness assumptions in a number of papers in 1973,'74,'75,'76 [24,25,26,29] .
210
SHAMIR in 1966,'67 [87,88] studied systems of singular integral equations in LP( ~+ ) , ~UBIN in 1971 [90] and ESKIN in 1967,'73 [104,30] applied the method of n matrixfactorization with respect to the nth Fouriertransform variable ~n in order to solve general boundary value problems for pseudodifferential operators being homogeneous of degree ~ + iB in ~n with additional potentials and/or + boundary conditions on ~ = R'n1 added In his book (1973) ~SKIN is also concerned with mixedtype boundary conditions on ~, nI so that boundary pseudodifferential arise. operators with piecewise continuous coefficients on ~,nI
A main part in the technique is played by the "Mellin transformation" and the "Mellin convolution". This enters in a natural way by studying certain operatoralgebras acting on LP(R+). To show this we are going to follow the lines of KREMER (1969) whose work strongly parallels CORDES'(1969). In order to build up the algebras he introduces for ~EL2(IR+) : (5.10) (5.11) (5.12) (5.13) S T : = SIR+ : = T]R+ defined by ( S ~ + ~ ) ( x ) : = ~Ti ! ~ i +xmlY)d~ y '
defined by (T~+m)(x) : = i
_ 1 ! k(xy)~(y)dy _2 ~ o 7 k(x+y)m(y)dy 1
Then the basis for the whole theory are the compactness relations given by the following [55], p. 4 & 11: Lemma 5.1 : (5.14)
(5.15)
Let S2
a(x)cC(~+) = I + T2
, i.
and
S,T,Kk,C k with
kcLl(~)
as above Then
[S,T]~O
e.
C~(L2 ( ~+ ))
lim a(x) = 0
[al,T] ~ 0 in case of
211 Then KREMERstudies, as does CORDES (loc. c i t . ) , the algebra (~, generated by S and T which is commutative, modulo the compact operators, with an i d e n t i t y . D e f i n i t i o n 5.1 : (5.22) ~(t) If mcL2(~+) then
: = l.i.m. ~ ~0
is called the "Mellin transform of I t has the property (5.23) (5.24) (S~o)(t) = tanh (Tm)(t) =
1,11
 ~'(t)
and
 i ~'~t) cosh ~ t 2 (cf. e.g. CORDES [11], p. 897!). The algebra ~ I proves to be isometric isomorphic to the function algebra C ( ~ ) ( ~ = ~ in this context ) where ~(t)EC(~) corresponds to K EC~1 with (K~)(t) = ~(t).~'(t). ~1 is called the "generalized Mellin convolutional algebra" since i t contains an algebra O~ of integral operators of a "quotient convolution" type
(525) : dy , x > o
which transforms
o mELl(xi/2; ~ + )
qcLl(xI/2;
~+
(cf. e.g. CORDES [11] , p. 897/88!),The general Mellin transform (Mm)(s) : = i x S  l ' m ( x ) dx transforms such a quotier~ convolution into an algebraic product: o
(5.26)
M(gem)(s ) = (Mg)(s).(Mm)(s)
(cf. e.g. TITCHMARSH's book, p. 304). CORDES & HERMAN in (1966) [15] introduced the algebra ~ generated by Ct I , the m u l t i p l i e r s a ( x ) . l EC(~'~) and ~ ( L 2 ( ~ + ) ) such that ~ / ~ is a commutative B*algebra and the space of i t s maximal ideals is homeomorphic to the Shilov boundary ~ ( ~ x ~ ) = : v ~ . I f denotes the Gelfand homomorphism of ~ / ~ o n t o C(~() then we have the symbols OA(x,t) ECOl,) given by aal(X,t ) = a(x) ~Kq(X,t) = ~ ( t ) on on 0 ~ x ~ ~ , t = ~ ~ ~ t ~ +~, x = 0 or
+~
and Os(X,t) and ~T(x,t) as in eqs. (5~23), (5,24). This has been generalized to the corresponding subalgebras of ~ ( L P ( ~ + ) ) , l<p<~, by GERLACH & KREMER in 1972,'73 [39,40] where only
212 t = tanh {~ [~ + i ( ~ i _~ ) ] }
~P)(x,t) changes.
Now, i f we apply these r e s u l t s we get Theorem 5.1 : Then (5.27) (Kom)(x) : : a(x)~(x) + b(x)(S~)(x) L2(R+) 0 iff (CORDES (1966), KREMER (1969)). Let a,bEC(R +) be given.
or, e q u i v a l e n t l y , i f f on ~+ and
be FredholmNoether and the c o e f f i c i e n t s to be normalized s. th. on ~... I f we define KI : = a l l + b l S then KI where bI = b(a2b2)  I = b (5.32)
In order to s i m p l i f y the w r i t i n g in what follows l e t us assume Ko = al + bS to a2(x)b2(x) = 1 aI = a(a2b2)  I = a is FredholmNoether a l i k e with f darg DE ~Ko(X't) have zero indices but are in general not and
KREMER's idea (th. 23, p. 51) [55] is to construct a twosided r e g u l a r i z e r L = I + blbT 2 by i n v e r t i n g the symbol OL(X,t). He then i n v e s t i g a t e s the WienerHopfalgebra 6(2 ators K , I k and the compact ones on L2(~+ ) ~2 = ~ 2 / ~ i s ator norm of ~ ( L 2 ( ~ + ) ) .
213 C ( ~ ) . The Gelfand theory then allows us very easily to characterize the Fredholm operators, that are the i n v e r t i b l e elements in ~ 2 ' of the type by the condition i + k(~) # 0
A
I + Kk on
L2(R+),
on
as we know already.
To t r e a t the compound integral equation on L2(~+) the algebra ~N. being gener~ ating by OC and 02 is constructed. Again ( X / ~ i s commutative and semisimple with I an i d e n t i t y . The space T~C(~J of maximal ideals is homeomorphic to a compact subset of J~, (cf. KREMER[55] (ths. 19,20) where ~: = ~ u{(t,~) : t~, ~ =0}
: = (m~)
Then the compound integral equation with constant coefficients (5.35) Km e (I +aKk+~SKk)m(x ) = f(x)CL2(IR+)
may be treated. There holds the Theorem 5.2 : (KREMER, th. 24, p. 54). Let m,BEC , k E L I ( ~ ) , Then Ke ~c ~, and ~(o) = o , I f the function G(~) : = I + (~+B'sign ~)~(~) @ 0 on ~ then (i) KE~(L2(~+)) (5.36) 5.37) (ii) (iii) ind K = v(K) =  ~ [arg G(~)] =
is a twosided regularizer to
A
k 1, k2ELI(~)
(5.38)
1 + k1,2(~) = [1 + (~B)k(C)] 1
as elements of the Wieneralgebra i ~ ( ~ ) , Now, similar to the argument by GERLACH [38] (cf. p. 27/28!) one is ready to discuss the general case of continuous coefficients a,b,c on R+: (5.39) (W~)(x) : = (a(x)l+b(x)S+c(X)Kk)~(x) = f ( x ) E L 2 ( ~ + )
making use of Rakovsh~ik's r e s u l t of (1963). So KREMERarrives at his main r e s u l t , viz. Theorem 5._33 i ([55], th. 26, p. 62). Let Let and a,b,cEC(~'~), k e L l ( ~ ) , al + b.S on T+~ be
a
i +(~B)k(~) # 0 on
214
(5.41)
wh~re (15.42) Then W is
1 + k ( 0 ) { ~  s . t a n h T  ~p(t) + B tanh ~t
p(t))
# 0
on
]Rt
p(t)
: = b2(=).[b2(=)
+ cosh 2 _~]1 .
T(L2(m+))
oo
C.rollary 5 . 1  
Let
a,b,c,k Ro
Then the second c o n d i t i o n f o r the symbol is a u t o m a t i c a l l y f u l f i l l e d a twosided r e g u l a r i z e r (5.43) where (5.44) (5.45) UI : = al  bS U2 : = I  b ( x ) . b ( o ) [ l  t a n h ~] Kq1 b(x)b(~) tanh
R : = U3U2UI
Kq2 .
(5.44)
where
K ~ c ~ 1 ; i = 1,2 ; and qi q1(t) : = [cosh 2 3_~ + b2(0)]  I (5.45 a) q2(t) : = [cosh 2 ~ + b2(~)]  I
(5.45 b)
ind K : v(K) = ~
(cf.
th. 1.4!) where a(O)+b(O) a o = arg a(O)b(O) = Co + 2~ mo defines mo,m~ a iCol, = arg Ic I < th. 2.7), Let a, b, c, k be as above but b(o) = b(~) = O, = c2~ m
I + ~K k
be c ~ ( L 2 R + ) ) .
V~O
Now, we shall sketch the theory developed by GOHBERG & KRUPNIK [45] f o r s i n g u l a r
215 Cauchytype integral equations involving piecewise continuous c o e f f i c i e n t s . While the whole theory has been developed f o r spaces LP(p;c), l<p<= , p(t) ~ O, p , plqeLl(F) f o r I / p + I / q = 1 and FC~ , a system of Ljapounov arcs and/or contours, we shall confine ourselves here to one closed contour Let rcC (or s = ~ ) and L2(F). Ne are following mainly along the lines of t h e i r paper in (1970) ~5 ]. PC(F) denote the algebra of a l l piecewise continuous functions a(to) = a(toO ) = lim a(t) and a ( t ) on F being l e f t continuous having l i m i t i n g values
from the r i g h t
lim a ( t ) where ~ and ~ shall denote that the point t t to t >t o is before or behind t o , respectively, in the sense of the orientation on F . G : = F ~ [ O , I ] shall denote the cylinder { ( t , u ) : tEF , O~u~1} and M(t,p)#~2x2(F), i.e. a 2x2complex matrix whose entries O l l ( t , ~ ) , ml2(L,u), m21(t,~), and
a(to+O ) =
~22(t,1~)~C(G) where ~12(t,0) = ~21(t,0) = ~12(t,1) = ~21(t,1) = 0 for tCF . Now, the algebra ~ of a l l such functionmatrices becomes a Balgebra by introducing the norm (5.48) l l M ( t , ~ i i : = max Sl(M(t,~)) where s~(M(t,~)) denotes the largest eigenvalue of H(t,~).M*(t,u) being
semipositive d e f i n i t e . They prove then the following theorems in several steps: Theorem 5.4 : ~(L2(F)) a,bEPC(F) (i) ( [ 4 5 ] , t h . 0.2, p. 194). Let OL(PC(F)) the ideal of compact operators on be the smalle~subalgebra of L2(F). Then ~ . This iso
(~t(PC(s))/~
morphism transforms the operator C~mA = a , l + bS F + V , Vc~, into the "symbol" aA(t'~)=~(t'~):= I(1~)c(t)+~,c(t+O) (5.49) ~[c(t+O)c(t)], where c(t) : = a ( t ) + b ( t ) , d(t) : = c ( t )  b ( t ) (5.50) ( i i ) Remark 5.1 : ~d(t)+(Z~#)d(t+O) / , ~[d(t+O)d(t)]l
ll~(t'~)ll
= Vc~]infIIA+V]I
(L2(r))
of integral
Since they show (th. 3.2) that for a matrixfunction M(t,~) e ~ with det M(t,~) ~ 0 on G ( M ( t , ~ ) )  I E ~ too, they can prove the following essential (th. 4.2, p. 199!):
216
Let on
AE~(PC(F))c~(L2(F)), G~.
Then
AE ~ ( L 2 ( ? ) )
or
e~_(L2(F))
iff
AE~(L2(F))
(5.52)
ind A = v(A) =  ~
Remarks : 5.2 : In t h e i r paper 1971 [46] they carry over t h e i r arguments to the m Bk spaces LP(p;?) , l<p<~, where p(t) : = k~ 1 ItCkl w i t h 1<6k<P1 and points Cker 5,3: fixed (cf. also t h e i r j o i n t book (1973)[47], chap. I X ! ) . s i n g u l a r i n t e g r a l operators number of
treated n o n  e l l i p t i c det~(t,~)
l i k e above in
v ~ ' P ( p ; r )  s p a c e s where
may have a f i n i t e
i s o l a t e d zeros of f i n i t e DORF's f o r closed curves with nonelliptic chapts, 5,6 !). 5.4:
symbols
GOHBERG& KRUPNIK discuss In sects. 6 and 7 of t h e i r book (1973) [47] the cases
of F = ]R and .IR+in LP(Po;? ), i< <=, and the weiaht f u n c t i o n Po(t ) ( l + x 2 ) ~12 N JXXkl 6k 1 < z N p ;I ,Bk+B<P1 or, p a r t i c u l a r l y in 7, the operator k=l k=l corresponding to the "twopart composite s i n g u l a r c o n v o l u t i o n a l equations" (cf. remark 5) of ~hap. 3 b e f o r e ! ) :
(5.53)
A = a.l+b.S]R
where
a =
a I c~ a2 e~
i n v a r i a n t operators x]R+
FlaFim and
Wa : = P+Wa IR~P+)
~j
P+=
.I ,
a(~) =
P~I~(~) denotes the class of a l l f u n c t i o n s on ]R of type n ^ z ak(~)k(~ ) where the ak(~) = c k + g k ( ~ ) E ~ ( ] R ) , the onedimensional k=1 Wieneralgebra, and the x k being c h a r a c t e r i s t i c f u n c t i o n s to i n t e r v a l s EkC IR
217
such t h a t
EkFIEJ = (~ and ~
bounded t r a n s l a t i o n To each f u n c t i o n
U Ek = IR. W is then the unique c o n t i n u a t i o n to a a k=1 i n v a r i a n t operator on L2(IR) ( c f , [24], chap. 2, th. 2 and 1!) he associates the f u n c t i o n f o r I~I < ~ and O~u~l for ~ =
a(~) E P ~ ( ] R )
{ a(~O)[lu]+a(~+O).u (5.55) a(2)(~,~) : = a(+~) [1~]+a(~) having a closed curve gular" if cC "u
aEP~(R) is
c I . . . . . c n. Then
W = P+FIaFP+E~,(L2(~+) a invertible if
E ~+(L2(~+~
I f t h i s c o n d i t i o n holds then
(5.57)
a(~O)[1gp(~)]+a(~+O).gp(~)
[a(+~) [ 1  g q ( ~ ) ] + a (  = )
r = p or gr(~) : _
and 5.7 :
ak(~ ) = c k + ~k(~ ) , g k ( X ) c L l ( ~ ) n L q ( ~ ) . He constructs an algebra of operators A = r s k=l s ~ j=l W a kj where akj E P~]~( ~R )
(5.59)
218
holds. Then
Os~l
A d e t a i l e d representation is given by him (1975) [26 ]. 5.8 : DUDU~AVA in (1974) [25] generalizes his theory to include m u l t i p l e  p a r t
S (Am)(x) : = j=1 a j ( x )  ( W b j . c j ( y ) m ) ( x ) bjeP?A)(~) as before, a j ( x ) p = 2 in and cj(x) ~ C(R), the closure of the L~functions are
)
on
, i . e. f o r L2(~).
L~(~)~Jorm, since a l l
The author constructs complicated 2 x 2symbol matrices along and calculates the index. The d e t a i l s are too lengthy to
the l i n e s of GOHBERG & KRUPNIK and formulates necessary and s u f f i c i e n t conditions A to be e~(L2(~)) be w r i t t e n down here! his paper (1976) [29] DUDU~AVA gives a d e t a i l e d account of the whole theory extending i t to the quarterplane case, permitting Sobolev spaces, and systems as w ~ l . 6. Convolutional i n t e g r a l equations on the 9uadrant A (iii) in Chap. 1 l e t us consider the f o l l o w i n g "WienerHopf
In accordance with
i n t e g r a l equation on the quadrant" (6.1) where (W++ m)(x) : = m(x)  f kcLI(R 2) and f are given, k(xy)m(y)dy = f ( x ) E L P ( ~ + ) 2 mcLP(~++) , x2 ~ 0 } . If sought, ~ + ^ d e n o t i n g ~(~) : = 1k(~) the f i r s t # 0 on ~2
quadrant c III0(~)
= {x = (Xl,X2)E~ 2 : xI ~ 0
then we may f a c t o r i z e i n t o four continuous functions which are holomorphically ex+ tendable into the four respective products of halfplanes H~i ~< H ~2 such t h a t (6.2) (6.3) ~(~) = ~++(~)~_+(~)~_~(~)~+_(~) where
a,(~) =
A
1 + ~+,+(~). = exp{Pt, +~ log ~(~)} . . . . are the Ftransformed projectors onto the four
P, : = FX 2 . I . F 1 quadrants of
219
r e s p e c t i v e l y . Denoting the halfplane WHOs by TD (W) and Tp (W) , r e s p e c t i v e l y , ~m where W : = l  k ~ is the twodimensional L 1convolution on u ~ L , we have the inverse by GOLDENSTEIN & GOHBERG ( c f . remark 2.1!) e x i s t i n g f o r (6.4) [TPr(W)]I = FI[~++~+_(~)] 1FP r FI[o_+~__(~)]IF [TPu(W)]I . Due to a r e s u l t by SIMONENKO (1967) [94] is compact on Lp( ~ 2 ) , l<p<~ , f o r quadrants E,Gc~ 2 aW(~) # 0 on ~2 as
lying opposite on the same diagonal we a r r i v e at the Theorem 6.1 : (STRANG (1970)[ 99 ]) Let kELI(R 2) and inf.~ I I  ~ ( ~ ) I > O. Then 2 the WHO W++EBm(LP(~++)), l<p<= , with ind W++ = 0 with ~ E ~ : t h e r e g u l a r i z e r (6.5) R = P++{[TPr(W)]I + [TPu(W)]I  WI}P++
The present author and F.O. SPECK [59] r e c e n t l y studied the WH integral equation with (6.6) kELI(~ 3) and
where G = G(~) : = {x = (Xl,X2,X3) : x l + i x 2 = re i ~ , o ~ _ _ ~ , r ~0, x3E~}= S ~ denotes a wedge with x3axis as i t s edge and with an opening angle equal to ~ . F i r s t they proved the Theorem 6.2 : Let iff ( [ 5 9 ] , th. 1). Let hE~ n Zc~ n , n ~ 2 , E ~ ( L 2 ( ~ n ) . Then be a c y l i n d r i c a l also region, i . e.
such t h a t f o r a l l xcZ
x + phEZ f o r a l l p ~ O.
A be t r a n s l a t i o n i n v a r i a n t
Tp(A) : = PzAI~9~(pz)~(L2(Z))
Remarks :
spaces and spaces with a weight function. 6.2 : The r e s u l t carries over to cones ~A(p~) : ~A(~) for p > 0 Fc~ n instead of and ~ c ~ n  { o } . Z admitting " d i l a t a t i o n
i n v a r i a n t " operators, e. g. p s e u d o  d i f f e r e n t i a l behaves l i k e Now, the f o l l o w i n g r e s u l t is true Theorem 6.3 : are equivalent: (i) (ii) WGE~(L2(G)) WGE ~(L2(G)) is i n v e r t i b l e and ( [ 5 9 ] , th. 2). Let
220
(iii) every operator W in the family defined by F1,2 ~( " " ' x 3 ) F I , 2 ' 1 ~,x3 X3E~, is i n v e r t i b l e as twodimenslonaL W O on the sector S~ with angle ~ at the H vertex. While in the case of G = ~n or n = ~+ , n ~ 2 , the e l l i p t i c i t y
i n f I~w(~)I > 0 ~wE}q@(~ n) is necessary and s u f f i c i e n t for the i n v e r t i b i l i t y of ~E~n ' , W operating on LP(~ n) or L P ( ~ ) , l<p<~, f o r general s e m i  i n f i n i t e GcRn not so much is known but GERLACH & LATZ proved 1977 Theorem 6.4 : ellipticity (6.7) for suitable Let A = l  k , , kELI(Rn), E 9 r Tp(A) : = PEAI~(pF) be a measurable subset in i . e. ~n con
to be i n v e r t i b l e on "~.(PE) ~ L2(E)
~ > 0 , is s u f f i c i e n t .
Cor011ary 6.1 : i t is a)
([59], Corollary 1). For the W O W of eq. (6.6) to be i n v e r t i b l e H G lk(~) be e l l i p t i c and b) s u f f i c i e n t that i t be point
necessary that
In order to admit strongly singular convolutional operators on quadrants depending even on the variables Xl,X 2 we introduce the theory of operators of " b i  l o c a l type" by PILIDI (1971) [66] which has a predecessor in SEELEY's paper (1965) [82] Chap.13~, in a sense (see also DOUGLAS & HOWE (1971)[23]!). D e f i n i t i o n 6.1 : Let ~1' ~2 denote two Bspaces of Lpfunctions on ~m and ~ n , respectively. Ajc ~ ( ~ j ) the operators of local type on Xj ~j the set of operators which are t r a n s l a t i o n i n v a r i a n t by BIB 2 c~(~j), "~j the subalgebra and ~ the BI and ideals of the Fredholm and compact operators on ~ B2 we have: a) A,BEA : = AI~ A2 b) , respectively. Then denoting
the topological (and algebraica]) tensor product of two Bspaces are called " i  e q u i v a l e n t " , A~B , i.ff ABE~I@A 2
c)
A,BCA are called " l o c a l l y Iequivalent at exists neighborhood such that ~ (Xo)C~ m
there
and ~uEC~(~m)
with
~L(x) ~ 1 on ~ (Xo)
(6.9)
<
221
d)
AEA is called "locally 1FredholmNoether at x~E~ m'' i f f there is a neighborhood 1~L(x~)c~m, a~Lcc ~~ ) ( "m as above, and R~,x~ , Rr,x~ E A such that R~ o a(o~Iz 12)~i ( ~ l l ~ 1 2 ) A R r , x ~ ~,x I ,~1 ~ I i ~ 12
Theorem 6.5 : (PILIDI (1971)[66]) : AEA = A1(~(1)A2(~2) is c~(JEI~)E2) i f f A is locally lFredholm for all x~c]R and is locally 2Fredholm for all x~E]R m n. Theorem 6.6 : (PILIDI (1971)[66]): Let A,BEA and A be locally lequivalent to B at xOER . Then A and B are at the same time locally lFredholm at x~ or m 1 not. These oheorems are applied to "bisingular Cauchytype integral equations on ~2 or ~ m ~ n , , and to WienerHopf integral equations on the quadrant. Theorem 6.7 : Let a (Xl,X2)EC((~)2), ~ = 0,1,2,12, and the bisingular Cauchytype operator be defined by (Lm)(Xl,X 2) : = ao(Xl,X2)~(Xl,X2)+al(Xl,X2).(Sl~)(Xl,X2) (6.11) where (6.12) (Sl~)(Xl'X2) : = ~ i f (Yl,X2)dY 1 Yl  Xl ~(xl,Y2)dY 2 (S2m)(Xl'X2) : = ~ i f ~2 denote the "partial Cauchy transforms" and (6.14) (S12~)(Xz,X2) : = (Sz(S2~))(Xz,X2) = ~ Y2  x2 +a2(xl,x2)(S2~)(x1,x2)+a12(Xl,X2)(S12m)(Xl,X2 )
(6.13)
Then the following statements are equivalent: LE~(L2(IR2)) (6.15 a) for all (6.15 b) ZlE~ I
[ao(Xl,Z2)+a2(xl,z2)sign C2]I+ a1(xl,z2)+a12(x1,z2)sign ~2 ] SI and fixed C2 = 1 are invertible onedimendional singular operators x 2 and x I , or the symbol of L :
222 (6.16) aL(Z1'Z2;~l'~2) : = ao(Z1'Z2)+al(Zl'Z2)sign ~1 + + a2(z1'z2)'sign ~2+a12(z1'z2 )sign ~I "sign ~2 is e l l i p t i c (6.17 a) and (6.17 b) [arg aL(Zl,Z2;~l,~2)] = = 0 for all Zl=~ In case of avEC(~2 ) z2c~ 2 ; ~i,~2 = 1 i. e. # 0 on I~2 ~ {  1 , 1 } 2 and a d d i t i o n a l l y = 0 for all ZlC~l ; ~i'~2 = I
Remarks : 6.3 :
6.4 : This theorem holds also for R I ~ ~2 replaced by Ljapounovcurves rl,r2cC and Bspaces LPl(pl;F1) and LP2(p2;F2) instead of L2(~1 ) and L2(~2 )" The coefficients a may even be piecewise continuous only or systems of equations with matrices a may be involved. (cf. mainly PILIDI & SAZANOV (1974)[67] and DUDU{AVA (1975) [27,28]. They calculated a twosided regularizer and the index of the operator L.
Corollary 6.2 : Let LEA = AI(LPl(pl;FI) ) ~ A 2 ( L p 2 ( p 2 ; ~ ) ) be a FredholmNoether operator. Let R1 and R2 denote I  and 2 partial regularizers, respectively. Then a regularizer (6.18) and the index R for L is given by
R : = R1 + R2  RILR2 L by
(6.19)
where
Remarks : 6.5 : Applying the twodimensional Ftransformation to the WienerHopf equation on the quadrant ~ +  or more general: studying fourpartcomposite equations
IR2
~r
~2
4
where ~ = ~++E$ ~ leads to Riemann boundary value problems for two complex variables in the four products H of halfspaces in C . This is then equivalent to an equation like (6.11) with operator L . The problems have been thoroughly discussed mainly by DUDU~AVA in 1976 [ 29] ' 2,3, in Sobolev spaces (Hs'P(~ 2+ ))m ' . mE~", + i. e. including systems. 6.6 : PILIDI & SAZANOV (1971,'74) [66,67] treated also operators A of the bisingular CMOtype where AI~XAxEQmA2(~ n) for all x ~ m and A2'~YBy~AI(~m)Qn for all yE~ n, where the Qm' Qn denote m u l t i p l i c a t i o n by homogeneous functions of degree zero Ec(~m{o}) and E c ( ~ n  { o } ) , respectively. 2 6.7 : KREHER(1976) [56] studied n o n  e l l i p t i c WienerHopf operators on ~++ where the symbol may have a f i n i t e number of zero lines of f i n i t e order in each of the two variables 7. ~i,~2.
Concluding Remarks
SHINBROT in 1964 [ 89], DEVINATZ & SHINBROT in 1969 [16], REEDER in 1973 [76], and PELLEGRINI in 1973 [65], j u s t to mention a few papers, were involved into the study of general WHOs on H i l b e r t spaces ~ ; Tp(A) i~(p ) where AE~(~) and P = P2E ~(~.) a projector. They derived c r i t e r i a for the i n v e r t i b i l i t y of Tp(A) on ~(P) and the connection to f a c t o r i z a t i o n . Concerning e l l i p t i c systems of singular integral operators having symbol matrices being (~+l)times continuously d i f f e r e n t i a b l e positively homogeneous functions of ~ : ( 6 1 . . . . ~n) , ~ > n , the i n v e r t i b i l i t y of the ~ corresponding WHOs on ~+ has been discussed by SHAMIR in 1966,'67 ~7,88] in w S ' P ( ~ )  s p a c e s . He derives a p r i o r i estimates and makes use of the f a c t o r i z a t i o n of matrices according to GOHBERG & KRE~N (1958)[43]. See also the paper by ~HUBIN in 1971 [90]! These problems have been displayed and treated by the Banach f i x e d point principle for strongly e l l i p t i c translation invariant A by HEISTER & SPECK (1977/78) [58]. They also discussed the more general problem of an "Npart conjposite WHO"
N
n
~JN : =
z Aj,Pj j=l
on
LP(~ n) , l<p<~,
AjE~(LP(IRn))
are i n v e r t i b l e and
with
I.
In t h i s paper one may find also a couple of examples of mathematical d i f f r a c t i o n theory leading to such operators (cf. [58], chap. 2!). The present research centers around the following questions, at least in the mind of the author: (1) Which are the necessary conditions for an AE~(3C) and a given continuous
projector P on ~ for Tp(A)_ _1~l~,p) to be invertible? The same question applies N to ~ AjPj to be i n v e r t i b l e on ~. j=l In order to allow xdependent symbols ~A(X,~) of generalized translation invariant operators or pseudodifferential operators one wants to know: (2) Which are the necessary conditions on ~A(~) EL=(~ n) such that A = FI~A(~)F ~EC(]R n) or is
is compact on
 or subspaces  so that
~(x).FI~(~)F
compact on ws'P(~ n) scIR , l<p<~, fixed? In the case of a positive answer one may obtain a Fredholm c r i t e r i o n for this more general class of operators. Some information concerning the l a s t question has been given recently by CORDES in 1975 [13] and by SPECK in (1976,'77') [97,98] generalizing the results by RAKOVSH#IK (loc. c i t . ) Most questions on mixed boundary value and transmission problems f o r e l l i p t i c p a r t i a l or pseudodifferential equations may be put into a generalized WienerHopf s e t t i n g , as the work by ESKIN, DIKANSKII, RABINOVI~, SHAMIR, SEELEY, BOUTET DE MONVEL and many others show, but the d e t a i l s concerning the question above s t i l l worked out. have to be
Prof. Dr. Erhard Meister Fachbereich Mathematik Technische Hochschule Darmstadt Schlo~gartenstr. 7 D 6100  Darmstadt
225
References 1 2 3 4 5 6 7 8 9 I0 II 12 13 14 15 16 17 18 19 20 21 Agranovi~, M.S. E l l i p t i c singular integrodifferential operators. Russ. Math. Surveys 20, 1121 (1965) Atkinson, F.V. The normal s o l u b i l i t y of linear equations in normed spaces (russ.). Hat. Sbornik N.S. 28 (70), 314 (1951) Bancuri, R.D. Integrodifferential equation on the halfaxis with kernels depending on the difference of arguments (russ.), Sakharth. SSR Mecn. Akad. Moambe 54, 1720 (1969) Betz, A. TragflUgeltheorie . Ber.u. Abhandlq. d. Wiss. Ges. f. Luftfahrt, Heft 2 (1920) Boutet deMon~el, L. Op~rateurs pseudo diff~rentiels analytiques et probl@mes aux limites elliptiques . Ann.lnst. Fourier 19, 163268 (1969) Boutet de Monvel, L. Boundary problems for pseudodifferential operators Acta Math.126, 1151 (1971) Calder~n, A ~ . Singular integrals , B u l l . Amer. Math, Soc. 72, 427465 (1966) Calder6n, A.P., Zygmund, A. On singular integrals. Amer. J. Math. 78 289309 (1956) ~ Calder~n, A.P., Zygmund, A. Singular integral operators and d i f f e r e n t i a l equations. Amer. J. Hath. 79, 901921 (1957) Colella, P., Cordes, H.O. ~ e C~algebr~ of the e l l i p t i c boundary problem. SFB 72, preprint no.160, Univ. Bonn 1977 Cordes, H.O. Pseudodifferential operators on a h a l f  l i n e . Journ. Math. Mech. 18, 893908 (1969) CordesTH.O. Lecture notes on Banach algebra methodsin partial d i f f e r e n t i a l equations. Univ. Lund 1970/71 Cordes, H.O. On compactness of commutators of multiplications and convolutions, and boundedness of pseudodifferenti~al operators. Journ. Funct.Anal. 18, 11513! (1975) Cordes, H.O. A global parametrix for pseudodifferential operators over ~n, with applications. SFB 72 preprint no.90, Univ. Bonn 1976 Cordes, H.O., Herman, E. Singular integral operators On a h a l f  l i n e . Proc. Nat. Acad. Sci. USA 56, 16681673 (1966) Devinatz, A., Shinbrot, M.General WienerHopf operators . Trans. Amer. Math. Soc. 145, 467494 (1969~ Dikanski~, ~ . Problems adjoint to e l l i p t i c pseudodifferential boundary value problems, Soviet Math. Dokl. 12, 15201525 (1971) Dikanski~, A.S Conjugate problems~f e l l i p t i c d i f f e r e n t i a l and pseudod i f f e r e n t i a l boundary value problems in a bounded domain. Math. USSRSbornik 20, 6783 (1973) ~ n i g , J. Zur Theorie einer Klasse e l l i p t i s c h e r singul~rer IntegroDifferentialoperatoren in Grund und Distributionenr~umen. Diss. Univ. TUbingen 1973 Donig, J. Das CauchyProblem fur eine parabolische Gleichung mit singul~ren Koeffizienten. Vortr. Tagg. Math. Meth. d. Phys., qberwolfach 1974 Donig, J. 'A priori'Absch~t~un~en fur eine Klasse elliptischer PseudoDifferentialoperatoren im Raum Lv(~"), Function theoretic methods for partial differential equations: Proc. Int. Symp., Darmstadt, 171191, Springer Lecture Notes 561, Berlin Heidelberg 1976 Donig, J. Ober ein Transmissionsproblem mit IntegroDifferentialbedingungen. manuscripta math. 23, 1 18 (1977) Douglas, R.G., Howe, R.On the Cmalgebra of Toeplitz operators on the quarterplane, Trans. Amer. Math. Soc. 158, 203217 (1971) Dudu~ava, R.V. WienerHopf integral operators with discontinuous symbols. Sov.~ath. DoKI. 14, 10011005 (1973) Duducava, R . V . O~convolution integral operators with discontinuous coefficients, Sov. Math. Do~l. 15, 13021306 (1974) Dudu~ava, R.V. On Wiener~opf integral operators (russ.).Hathem. Nachr. 65, 5982 (1975)
22 23 24 25 26
226 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Duduc~ava,R.V. On bisingular integral operatorsandconvolution operators on a quadrant, Sov. Hath. Dokl. 16, 330334 (1975) Dudu~ava, R.V. Bisingular integral operators and boundary value problems of the theory of analytic functions in spaces of generalized functions. Sov. rlath. Dohl. 15, 13241328 (1975) V Dud ucava, R.W Convolution integral operators on a quadrant with discontinuous sj~nbols. Math. USSR Izvestija I0, 371392 (1976) Eskin, G.I. Boundary value problems a~nd the parametrix for systems of e l l i p t i c pseudodifferential equations. Trans. Hoscow Hath. Soc. 28, 74115(1973) Eskin, G. Boundary problems for e l l i p t i c pseudodifferentiaToperators (russ.). 'Nauka', Hoscow 1973 Feldman, I.A. An equation of radiation energy transfer and WienerHopf operator equations (russ.). Functional. Anal. i. Prilo~en.5, I06I08 (1971) Feldman, I.A. On an iteration method for the equation oTradiant energy transfer, Sov.Hath.Do~l. 12, 10341038 (1971) Feldman, I.A. Operator ~J~nerHopf equations and their applications to the transport equation (russ.). Matem. Issled. 6,115132 (1971) Friedman, A. Partial d i f f e r e n t i a l equations. Holt, Rinehart & Winston, New York 1969 Friedrichs, K.O. Pseudodifferential operators, an introduction. Lecture Notes, Courant Inst. Hath. Sci, New York Univ. 1970 Gakhov, F.D. Boundaryvalue problems. Pergamon Press, Oxford et al. 1966 Gerlach, E. Z ~ Theorie einer ~ von IntegroDifferentialgleichungen. Diss. TU Berlin 1969 Gerlach, E., Kremer, M. Singul~re Integraloperatoren in LPR~umen. Zeitschr. Angew. rlath. Mech. 53, T 158159 (1973) Gerlach, E., Kremer,M. Singul~re Integraloperatoren in LPR~umen. Math. Ann. 204, 285304 (1973) Gerlach, E., Latz, N. Zur Spektraltheorie bei Faltungsoperatoren. Z.Angew. Math. Hech. 57, T 231232 (1977) Gohberg, I.ZT., Kre~n, M.G. The basic propositions on defect numbers, root numbers and indices of linear operators, Uspehi Mat. Nauk 12 (74), 44118 (1957) (russ.). AMS Transl. (2) 13, 185264 (1960) Gohberg, I.Z., Kre~n,H.G. Systemsof integral equations on a h a l f  l i n e with kernels depending on the difference of arguments. AHS Transl.(2) I_44, 217287 (1960) Gohberg, I . Z . , Krupnik, N.A. The spectrum of onedimensional singular integral o~erators with piecewise continuous coefficients. Matem. Issled. 3 1630 (1968) (russ.). AMS transl., ser. 2, 103, 181193 (1974) Gohberg I.Z., Krupnik, ~.A. Algebra g e n e r ~ d , b y onedlmens~onal singular integral operators with piecewise continuous coefficients. Funct. Anal, and its Applic. 4, 193201 (1970) Gohberg I.Z., Krupnik, N . A . Singular integral operators with piecewise continuous coefficients and their s~bols. Math. USSR Izvestija 5, 955979 (1971) Gohber~, I.Z., Krupnik, N . A . Introduction to the theory of onedimensional singular integral operators (russ.). ~ t i i n c i e , Kishinjow 1973. . . . . . . . G~Idenstein, L.S.,Gohberg, I.Z. On a multidimensional integral equation on a halfspace whose kernel is a function of the difference of the arguments and on a discrete analogue of this equation. Soy. Hath.Dokl.~,173176 (1960) GrabmUller, H. On linear theory of heat conduction in materials with memory, Existence and uniqueness theorems for the final value problem o Proc, Roy, Soc.Edinburgh 76A, II0137 (1976) GrabmUller, H. Asymptotic behavior of solutions of linear integrodifferen~ ial equations. J. Integral Equs. and Op.Theory l, 2856 (1978) H~rmander, L. Estimates for translation invar~ant operators in LPspaces. Acta Hath. I04,93140 (1960) Kohn, J.J.,N{renberg, L. An algebra of pseudodifferential operators. Comm.Pure Appl.Math. 18, 269305 (1965) Koppelman, W., Pincus~J.D. Spectral representations for f i n i t e Hilbert transformations, Math. Zeitschr. 71, 399407 (1959)
227 54 55 56 57 58 Kre~n, M.G. Integral equations on a h a l f  l i n e with a kernel depending upon the difference of the arguments. AMS Transl. 22, 163288 (1962) Kremer, M. Ober eine Klasse singul~rer Integ~Igleichungen vom Faltungstyp. DisS. TU Berlin 1969 Kremer, M. Ober eine Algebra 'nicht normaler' WienerHopfOperatoren I & II. Math. Ann. 220, 7786 & 8795 (1976) KumanoGo, H. Pseudodifferential operators ' (japan.), lwanamiShoten Publish., Tokyo 1974 Meister, E., Speck, F.O. Somemultidimensional WienerHopf equations with applications. FB Hathem., TH Darmstadt, preprint Nr. 373 (1977); to appear in Proc. 2nd Sympos. on Trends in Appl. Pure Math. Mech. Kozubnik, Sept. 1977 Pitman Monographs, vol. 2 (1978) Meister, E., Speck, F.O. WienerHopf operators onthreedimensional wedgeshaped regions, FB Mathematik, TH Darmstadt, preprint Nr. 389 (1978)~ to appear in Applicable Analysis. Michlin, S.G. Singular integral equations, Uspehi mat. Nauk (N~,) 3, 29112 (1948) (russ.),~IS transl. 24, 84198 (1950) Michlin, S.G. Multidimensional sTngular integrals and integral equations. Pergamon Press, Oxfor~ et al. 1965 Mushkelishvili, N.I. Singular integral equations. Noordhoff, Groningen 1953 Nickel, K. L~sung eines Integr'algleichUngssystems aus der TragflUgeltheorie, Math. Zeitschr. 54, 8196 (1951) Peetre, J. M i x ~ problems for higher order e l l i p t i c equations in two variables I. Annali di Scuola norm. Pisa (3) 15, 337353 (1961) Pellegrini, V. General WienerHopf operators and the numerical range of an operator. Proc. Amer. Math. Soc. 38, 141  146 (1973) P i l i d i , V.S. On multidimensional~isingular operators, Sov. Math. Dokl.l_~2, 17231726 (1971) P i l i d i , V.S., Sazanov, L.I. A priori estimates for characteristic bisingular integral operators. Sov. Math. Dokl. 15, 10641067 (1974) Pr~ssdorf, S. Operators admitting unfunded regularization (russ.). Vestn. Leningrad Univ. 20, 5967 (1965) Pr~ssdorf, S. Ei~imensionale singul~re Integralgleichungen und Faltungs91eichungen ni%ht normalen Typs i n lokalkonvexen R~umen~.~abil.schrif{,KarlMarxS'tadt (1967) " Pr~ssdorf, S. Zur Theorie der Faltungsgleichungen nichtnormalen Typs. Math. Nachr. 4_22, I03131 (1969) Pr~ssdorf, S. Einige Klassen singul~rer Gleichungen. Birkh~user Verlag, BaselStuttgart 1974. . . . . . . . . . . . . Rabinovi~, V.S. Pseudodifferential equations in unbounded regions with conical structure at i n f i n i t y . ~lath. USSRSbornik 9, 7392 (1969) Rabinovi~, V.S. Pseudodifferential equations in unbounded megi'ons. Soy. Math. Dokl. 12, 452456 (1971) Rabinovi~, V.S. Pseudodifferential operators on a class of noncompact manifolds. Math. USSRSbornik 18, 4559 (1972) Rakovsh~ik, L.S. On the theormof integral equations of convolution type (russ.). Uspehi Matem.Nauk 18, 171178 (1963) Reeder, J. On the i n v e r t i b i ~ t y of general WienerHopf operators. Proc.Amer. Math. Soc. 27, 7276 (1971) ~ahbagjan, R~.L. Convolution equations in a halfspace, AMS Translat.,Ser. 2, 75, I17148 (1968) Samko, S.G The general singular equation in the exceptional case. Diff. Equat. l , 867874 (1965) SchUppeT, B. Regularisierung singul~rer nicht e l l i p t i s c h e r Integralgleichungen mit unstetigen Koeffizienten, Diss. Univ. MUnchen 1973 SchUppel, B. Regularisierung singul~rer Integralgleichungen vom nichtnormalen Typ mit stUckweise stetigen Koeffizienteno 'Function theoretic methods for partial d i f f e r e n t i a l equations': Proc. Int. Symp., Darmstadt; 4 3 0  T , Springer Lecture Notes 561, BerlinHeidelberg 1976
59 ( 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
228 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 I00 I01 102 103 104 105 106 107 108 Schwartz, J.T. Someresults on the spectra and spectral resolutions of a class of singular integral operators. Comm. Pure Appl. ~,lath. 15, 7590 (1962) Seeley, R.T. I n t e g r o  d i f f e r e n t i a l equations on vector bundles. Trans. Amer. Math. Soc. 117, 167204 (1965) Seeley, R.T. Topics in pseudodifferential operators. PseudoDiff.Operators (C.l.M.E.,Stresa 1968). Edizioni Cremonese, Roma, 169305 (1969) ~ir, E. Evaluation dansW~,Ppour des probl~mes au l i m i t e s e l l i p t i q u e s mixtes dans le plan. C.R. Acad. Sci. Paris 254~ 36213623 (1962) Shamir, E. Mixed boundary value problems for e l l i p t i c equations in the plane. The Lp theory, Annali Scuola Norm.Sup. Pisa (3) !7, I17139 {1963) Shamir, E. Reduced H i l b e r t transforms and singular in~~gral equations. Journ. Analyse Hath. (Jerusalem)f2, 277304 (1964) Shamir, E. ~JienerHopf type pr~lems for e l l i p t i c systems of singular i n t e gral equations. Bull.Amer.Math.Soc 72, 501503 (1966) Shamir, E. E l l i p t i c systems of singular integral operators,l. The h a l f space case. Trans. Amer. Hath. Soc. ]_~_7_, 107124 (1967) Shinbrot, M. On singular integral operators. Journ. Math. Mech. I_33, ~95406 (1964) Subin, M.A. Factorization of matrices depending on a parameter and e l l i p t i c equations in a halfspace. Math. USSR Sbo~nik 14,6584 (1971) Simonenko, I.B. A new general method of investigat~nnq l i n e a r operator equations of the type of singular integral equations. Sov. Math. Dokl. 5, 13231326 (1964) Simonenko, I.B. A new general method of investigating linear integral operatorequations of the type of singular equations. (russ.). Izvest. Akad., Nauk, Ser. Matem., 29 567586 (1965) Simonenko, I.B. Convolution operators in cones. Sov. Math. Dokl. 8, 13201323 (1967) S~monenko, I.B. Operators of convolution type in cones. Math. USSR Sbornik 3, 279293 (1967) S~hngen, H. Zur Theorie der endlichen HilbertTransformation. Math. Z e i t schr. ~ 3151 (1954) Speck, F.O. Ober verallgemeinerte Faltungsoperatoren und eine Klasse von Integrodifferentialgleichungen. Dissert. TH Darmstadt 1974 Speck, F.O. Ober verallgemeinerte Faltungsoperatoren und ihre Symbole. Function theoretic methods for p a r t i a l d i f f e r e n t i a l equations: Proc. I n t . S~p., Darmstadt;459471, Springer L e ~ r e Notes 561'~, Berl~nJHeidelberg 1976 Speck, F.O. Eine Erweiterung des Satzes von Rak~vsh~ik und ihre Anwendung in der SimonenkoTheorie. Math. Ann. 228, 93100 (1977) Strang, G. Toeplitz operators in a quarterplane. Bull. Amer. Math. Soc. 76, 13031307 (1970) Talenti, G. Sulle equazioni i n t e g r a l i di WienerHopf, Boil. Un. Mat. I t a l . ~, Suppl. fasc. ~, 18118 (1973) 2nd Titchmarsh, E.C. Theory of Fourier I n t e g r a l s . e d . , Clarendon Press., Oxford 19~8 Tricomi, F. On the f i n i t e Hilberttransformation. Quart. J. Math. (2), 2 199211 (1951) , Vishik, M . I . , Eskin, G.I. Equations in convolutions in a bounded region. Russ. Math. Surveys 20, 85151 (1965) Vishik, M . I . , ~skin~G.l. Normally solvable problems for e l l i p t i c systems of equations in convolutions (russ.). Mat. Sbornik 74, (116), 326356(1967) Vishik, M . I . , Eskin, G.I. E l l i p t i c equations in c o ~ o l u t i o n in a bounded domain and t h e i r applications. Russ. Math. Surveys 22, 1375 (1967) Widom, H. Singular integral equations in Lp. Trans'. Amer. Math. Soc. 9_7_7, 131160 (1960) Wiener, N., Hopf, E. Ober eine Klasse singul~rer Integralgleichungen. Sitzg.ber. Preuss. Akad. Wiss.; Phys.Math. KI.3032, 696706 (1931) Zabreyko, p.p. et a l . Integral Equations  a reference t e x t . Noordhoff Intern. Publ., Leyden 197~ . . . . . . .
I.
Introduction. The most widely used method of solving boundary value problems for linear
elliptic partial differential equations is the classic separation of variables technique. Usually this approach leads to the study of eigenvalue problems
for linear ordinary differential equations containing a single spectral parameter, namely the separation constant. Indeed it is fair to say that this is perhaps
the prime motivation for the study of spectral problems associated with linear ordinary differential expressions. In many cases, however, when the separation
of variables method is effected it is found that the separation Constants cannot themselves be decoupled and several may appear as spectral parameters in each of the attendant ordinary differential equations. A classic example of this is the problem governing the vibrations of an elliptic membrane with fixed boundary. An application of the separation of
variables technique leads to the study of solutions of the following pair of linked eigenvalue problems for the Mathieu's equations.
d2yl
dx~ + (Xl + x2csh2xl)y I = 0, 0 ~ x I ~ ~, Yl(O)  regular, yl(~) = 0,
d2y2
dx~ + (%1  %2 cos2x2)Y 2 = O, 0 ~ x2 N 2~, Y2(X2)  periodic of period ~ or 2~. In this example %1 and %2 are the spectral parameters. Another example, recently App
230
d2y 1
dx]
{%1  %2 k2 sn2xl}yl
= 0,
K < xI < K,
x2
K + 2i~,
= 0.
Once again %1' %2 are spectral parameters and k, the modulus of the Jacobian elliptic function s n ~ is related to the semiapex angle ~ of the sector by k = sin ~. One can cite a wide range of problems leading to linked systems of ordinary differential equations containing 2, 3 or more spectral parameters. eigen
value problems gave way to some extent to the study of the "special functions" generated by such problems. Thus in the 1930~s and 40's there was much interest ellipsoidal wave functions,
in the properties of Mathieu functions, Lame functions, Heun functions and the like.
undoubtedly Arthur Erd~lyi who not only advanced our knowledge of these functions to the degree that we know them today, but also gave an encyclopaedic account of them in the well known volumes on higher transcendental arose from the Bateman manuscript project. functions [12] which
actively studied today and apart from their obvious interest to the mathematical physicist they do arise in other areas as well. Let us consider for the moment
one example which leads to an as yet unsolved problem. The problem we have in mind is of importance in the study of quasiconformal mappings[5] and may be described as follows. In Lam~'s equation let %2 = I
231
uniformisation theory asserts that for each value of the modulus k and consequently for each fundamental pair of periods of the Jacobian elliptic function sn x, is precisely one value h such that there
n_l (C
n2
L or
a half plane
J
In general
However the actual value of h is known in only a few special cases. we ask; for which values of h is
n~
(C  periods of snx)
n2
simply covered 1 ?
L Jordan domain
Clearly
V = ~ + m (m an integer) there are results due to Halphen LI4j, Ince [16J Indeed Ince [16j observed that when k 2 = ~ 2
some of the eigenvalues of Lam~'s equation giving rise to periodic solutions of periods 8K, 8iK' were rational and conjectured that this may be true for infinitely many values of v = ~ + m.
l ! 1
= ~ + m, %1 = ~ ~(~ + l) and k 2 = ~.
Lam~'s
nl'
~2 = ( s n x
d n x / k ) 1 / 2 2FI ( ~]m , ~ m1
then there are infinitely many rational values of %1 each giving rise to solutions of periods 8K, 8iK'. to the problems above. An important feature of the multiparameter eigenvalue problems which arise in mathematical physics is that the differential equations often have either singly or doubly periodic coefficients and it is to such systems that we now turn. 2. The Problem Consider the linked multiparameter system of periodic ordinary linear differential equations of the form There must be a large number of results of this kind relevant
232
n ~ %sars(Xr)Yr(Xr ) = O, s=l
r = 1. . . . , n. (2.1)
Here xr [0, ~0rj, (0 < 0o < ~) and the coefficients ars, qr are continuous r real valued functions with period ~r" determinant det{ars(Xr)} n r,s=! > O. We make the basic assumption that the
(2.2)
r = I, ..., n.
applied, see for example [20] or [4J. In addition to (2.2) we impose periodic boundary conditions of the form
(P)
(2.3)
or semiperiodic boundary conditions of the form (SP) Yr(0) = yr(~r), (2.4) Yr '(0)
= y~(~r ), r = l, ..., n.
As is now standard we say that the ntuple of necessarily real numbers % = (%1 . . . . . %n ) is an eigenvalue of the system (2.1, 2.2, 2.3) (or 2.4) if this r = l. . . . . n, for this
corresponding eigenfunction. The study of the above system and similar SturmLiouville problems has captured the interest of several authors recently. Indeed such problems may be The book of Atkinson
[4] and the monograph of Sleeman [20J are suitable references to this and other matters. From these works it may be readily verified that eigenfunctions and
eigenvalues do exist for the above problems and that the eigenfunctions form a complete orthonormal set in the weighted Hilbert space of Lebesque measurable functions on
233
n x r=l
E0, ~] ~
[0, ~r ],
compact intervals
[0, ~r j,
r = I, ..., n.
The
]lull2 = f
[0,~]
[ul 2 det{ars}dX
1 "'"
dx
n"
(2.5)
section we show how a classical argument well known in the the existence of eigenvalues and eigen
the eigenvalues
interlacing parameter
thus generalising
periodic case.
of eigenfunctions.
Recently Browne [8j has studied the interlacing problem and the stability of some
Browne and Sleeman [9j have obtained some results concerning eigenfunctions. notation. Let o = (oi, ..., ~n ) ~n, Or = 0 or I To describe
We
of type o applied
mean that periodic conditions are to be applied to the rth equation if O semiperiodic mnltiparameter conditions conditions eigenvlaue if o r = I.
= 0 and
(P) or (SP)
n}, r = I, ..., n, is a
This multiindex provides an ordering of the eigenvalues. spectra corresponding to the 2n different types
234
Consider
vectors
a, b c ~ ~
~n
r = I, ..., n.
i 0 ra r s ( X r ) l Y r ( X r )
12dXr,
r, s
= l, ..., n.
Next introduce
the set C c ~ n
defined by
C = {a e i~n I V(y)a < 0 for some Yr L2(O' Cr)' IIYrll = , Clearly C defines the 2n possible a cone and, as demonstrated by Binding
r =
l. . . . .
n}.
and Browne
[6], each of
Finally we define
i = {~~(~) l imultiindex;
~ ~
= 0 or I,
r = 1. . . . .
n}. periodic
Thus I is the collection of all eigenvalues or semiperiodic. The m a i n interlacing theorem of Browne
with 0 N e N (l,, i
(~~(~)
contains (i) (ii) any eigenvalue
(~~ ~(~)
c)
~,
J
X~(~) for w h i c h i < j < i + e and O r = Tr except that if er = then ~ r m a y be 0 or l, ; and (ir, T r) is (even, 0) or (odd, )
To illustrate
Let x l, x 2 e [O, lj and take Sl, s2, ql' q2 to be real valued functions on EO, IJ with p e r i o d one. system is
q2(x2)Y2(X2 )
+ (~'1
%2)s2(x2)Y2 = O.
The definiteness
condition
(2.2) is satisfied
since
det
Sl(X 1) S2(X 2)
Sl(X 1) s2(x 2)
is given by ~2
= 2Sl(Xl)S2(x 2) > O.
the cone
C = {(a,b) ~
I a ~ b ~ a}
and the spectral diagram is shown in figure I. If o = (0,0) the eigenvalues i ~ are marked and denoted by %~. If o = (0,1) the eigenvalues are marked o and i ~ ~ i denoted by ~~. If o = (I,0) the eigenvalues are marked m and denoted by ~~.
~ ~ i ~
In order to treat the stability question we must first define what we mean by stability as applied to the multiparameter problem.
Definition (i)
2.1 A point % ~ n is said to be a point of stability for the system (2.1) (~,~), r = I, ..., n. for the
stability
(~,~).
(iii) A point % ~ n which does not satisfy either for the system (2.1). (i) or (ii) is said to
be a point of instability
236
o~
O~
~0 ~
",.
/'
r s 1
e\ % % 1
i
i
>,
n
s I
p i z i
r i
Stability regions for Example I. Figure I. We now require to modify our definition of a cone as follows: multiindex i = (i, Given a
..., in) we define the cone C(i) to be the collection of all L 2 (0, ~ r ) such that
287
[V(f)a]r; the rth element of the column vector V(f)a, satisfies [V(f)a] r~ ~ 0 0 if ir is even, if i r is odd. Finally we denote Our
by S the set of all points of conditional stability of the system (2.1). stability result can now be stated as
i n {X~(1)  C(i)}],
(2.6)
where
I = (1,,
l)
This result is illustrated in the case of the above example by the shaded regions of figure I. For this example we have equality holding in (2.6). However
in [9J we give an example involving Mathieu's equation for which the inclusion in (2.6) is strict.
3.
Existence of Eisenvalues Here we apply a classic approach [I0] to the system (2.1) (2.2) (2.3) or (2.4)
to obtain some information regarding the existence of eigenvalues. Let ~r(Xr; X), ~r(Xr; X), r = ]. . . . . (2.1) satisfying the initial conditions ~r(0; ~) = I, ~'(0; r r = I, ..., n. The general solution to (2.1) can be expressed as a linear combination of the functions ~r and ~r" That is (3.2) ~) = 0, ~r(0; X) = 0, ~'(0; r ~) = i, (3.1) n be linearly independent solutions of
It is well known that a necessary and sufficient condition for the existence of twolinearly independent solutions each satisfying the periodic boundary condition (P) is
238
~r(mr; X) = I, Cr(mr; X)
=
~r(~r; X) = 0,
(3.3)
0,
~'(~
r
X)
1.
Similarly a necessary and sufficient condition for the existence of twolinearly independent solutions each satisfying the semiperiodic boundary condition (S  P) is ~r(~r; %) = I
T
~r(Wr; %) = 0,
!
(3.4)
~r(~r; X) = 0,
~r(~r; X) = I.
In general of course a necessary and sufficient condition for (2.1) to have a solution satisfying (P) or (S P) is that Dr(h) = ~r(Wr; X) + ~r(~r; X) = 2,
r =
(3.5)
~
...,
n.
The problem of existence of eigenvalues solvability of the system (3.5). method we find Dr(X) ~X
S
r r
2
X) +
JO {~r(T; %)~r(~r;
X)  ~r(~r; %)]
(3.6)
X)}ars(T)dT,
Now if D (%) = i2 r
det{ars} > 0 it follows from the inverse function theorem that (3.5) is uniquely solvable provided the eigenvalues are simple. If some of the eigenvalues are
not simple, in the sense that for some range of r, (3.3) and or (3.4) holds then we must work with the n x n derivatives of D r Hessian matrix constructed from the second partial
and make use of the inverse function theorem again. stability and interlacing theorems
239
of these results.
appear complicated.
In the following section we outline a different approach to the study of periodic multiparameter calculus of variations. eigenvalue problems. This approach is based on the
4,
The Variational Approach In this section we consider the case of twoparameters (n = 2) and study
the eigenvalue problem defined by 2 2 d Yr dx 2 + qr(Xr)Yr + s= ~s ars(Xr)Yr = 0, I~ r r = 1,2, x r e [0, ~ J r yr(mr) = Yr(0)exp i ~ t r, y$(Wr) = y$(O)exp i ~ t r, I < t ~ I, r r = 1,2, together with the definiteness condition (2.2). (4.2)
(4.1)
For this problem the existence of a countably infinite set of real eigenvalues can be established either from [20] or by the method outlined in the previous section. If we consider the eigenfunctions as being periodically functions the boundary
extended to the whole of ~2 as continuously differentiable conditions (4.2) may be rewritten in the form Yr(Xr + ~r ) = Yr(Xr)eXp i~ tr, r = 1,2. In addition to the condition that al2(X I) < 0 a21(x 2) > 0 on [0, ~i j, on [0, ~2 j.
(4.3)
(4.4)
This can always be arranged by a suitable scaling or affine transformation applied to the parameters %1' X2' As is well known [20] the eigenvalues and eigenfunctions (4.2) are simultaneous eigenvalues and eigenfunctions of the system (4.1)
240
~2y + a22(x 2)
~x 2
~x!
+ La12(x])q2(x 2)  a22(x2)ql(Xl)~Y
= % det{a
1
rs
}Y
(4.5) (4.6
Y(x + ~r ) = Y ( x ) e x p i ~ tr, r = 1,2. 82y ~2y al1(x;) ~  a21(x 2) ~  [all(Xl)q2(x 2)  a21(x2)ql(Xl)]Y ~x 2 ~x I = %2det{ars}Y, together with the boundary condition (4.6). defined as
~l = (el' 0), ~2 = (0, w2).
It should be noted that because of the assumed positivity conditions on a12 and a21 the left hand side of (4.5) is elliptic and it is to this equation that
most of our remarks are addressed. Let the eigenvalues of (4.5) (4.6) be denoted by An(t), (t = (t I, t2)) and It is readily
proved that the An(t) are real and form a countably infinite set with ~ as the only limit point. They may be ordered according to multiplicity as Ao(t) e Al(t) e A2(t) e .... and the corresponding eigenfunctions are orthonormal in the sense of (2.5). Notfce that since the eigenvalues %l(tl) of the given problem form a subset of the A (t) they exhibit a similar ordering to (4.7). n ~ theorem. Theorem 4.1. For each j = 1,2 let e x p ( i ~ t r of unity where I < tr. < I. J in any order. th roots ) (r~a = 0, I .... , k~  I) be the kj J Let ~R (I  R < klk 2) denote the pair (trl, tr2) < We also have the completeness (4.7)
241
1 ~ R ~ klk 2 is a complete set of eigenfunctions for the periodic problem over the rectangle
that the eigenfunctions of the given problem defined by (4.1) (4.2) are complete in the same sense. Now let
Q(Xl , x2) = a l 2 ( X l ) q 2 ( x 2 )
 a22(x2)ql(Xl),
(4.8)
and let F be the set of complex valued functions f(x) which are continuous in A ~ [0, ~i j x [0, w 2] and have continuous first order partial derivatives in A.
D(u,v) = fA{a22(x2) ~x I ~x 1 ~u ~V
for all u,v ~ F. This Dirichlet integral forms the basis for a variational approach to the multiparameter periodic problem. It is discussed, at least for SchrSdinger
operators, in Eastham [I] and its application to multiparameter SturmLiouville problems is considered in [19, 20]. Suppose we impose Dirichlet conditions on the boundary of A, i.e. Y(x) = 0 on ~A,
then for the associated Dirichlet problem we have corresponding eigenfunctions and eigenvalues which may be denoted by @n(X; 6n) and 6n respectively. In a
similar way, if we impose Neumannconditions on ~A then the associated Neumann problem will give rise to corresponding eigenfunctions ~(x; Nn ) and eigenvalues ~.
242
Now, arguing as in Eastham [II, p 101] we have the interlacing theorem Theorem 4.2. For n e 0 and all t ~n ~ An(i) ~ Nn" This simple interlacing result may be employed to give a further interlacing (4.10)
i
theorem for the twoparameter problem as follows.
i
~
value for (4.1) (4.2) then since the %7(t) form a subset of the An(i) theorem (4.2) gives, in an obvious notation, the result
for all t.
~ i
Substituting %7(t)into the first member of the system (4.1) we have the classic one parameter problem.
+ X2al2(Xl)Yl = O,
(4.12)
Yl(X! + e l) = Y l ( X l ) e x p ( i g t l ) ,
where of course al2 < 0. If we let the eigenvalues associated with (4.12) and the conditions
yl(wl) = yl(0) = 0,
i be denoted by 62 and those associated with (4.12) and the conditions y~(~l ) = y~(0) = O, i be denoted by n2 then again appealing to Ell, p 39] we have, for those eigenvalues
i
%~(t) which are simultaneous eigenvalues of (4.1) with r = 2, i i i
(4.13)
These arguments may be summarised in Theorem 4.3. Let the eigenvalues for the Dirichlet problem over A for (4.5) be denoted by
243
i i ~~ and the eigenvalues for the associated Neumann problem be denoted by n~, i i i where i = (i, i2) is a multiindex. If k~(t) = (%~(t), %2(t)) is an eigenvalue for the twoparameter tperiodic problem then i i i 8~ ~ k~(t)~ ~ ~n ~. i For this value of k~(~) we also have i i i
where B2, 62 are the Neum~nn and Dirichlet eigenvalues for the one parameter equation (4.12). There are a number of results to be obtained via the variational approach all of which extend in one way or another the results known for periodic Schr~dinger equations and promise interesting applications to the multiparameter periodic problem. We close this section with the remark that hidden in the
variational approach outlined above is the fact that the partial differential operator appearing in (4.5) is elliptic. For n ~ 3 we cannot always arrange for
any of the associated partial differential operators to be elliptic in general. In this case the arguments have to be modified and for details at least in the abstract case we refer to the monograph [20].
5.
The Alsebraie Approach In the study of periodic differential equations some useful insights are to
be gained by the study of various algebraic forms of the associated periodic differential equations. There is a strong connection here with the study of
Monodromic groups, the famous Riemann problem arising from Hilbert's 21st problem (see the article by Nicholas Katz in [7]) and also the study of quasiconformal mappings, the subject raised in the first section of this paper. Here we outline some of the results which are to be obtained in this setting and in particular the case of differential equations, like Lam~'s equation, which have doublyperiodic coefficients. [17, I, 2] as suitable references. when For a background to this work we cite
244
distinguished by having only two finite singularities whereas a doublyperiodic equation has three. The basic concept is that of a solution which is multiplicative path in the complex plane. for a given
(5.1)
(not necessarily regular) of (5.1), i.e. points where p(t), q(t) or both are singular, be denoted by tl, t2, ..., tn, ~. If to is an ordinary point of (5.1)
i
then we denote by Fi a simple closed path from to encircling the singularity t. oncepositively (clockwise) and enclosing no other singularity of (5.1). We
further denote by rij the path consisting of order and similarly for rij k etc. in the negative (anticlockwise)
ri, rj
Observe that
tl, ...
FI2 " "n is effectively a path making a negative circuit about infinity.
FI2 ... n is equivalent to the circuit r .
Since p(t), q(t) are rational functions of t the singularities of equation (5.1) are isolated. Hence if y(t) is a solution valid in a neighbonrhood of the along ri back to the neighbour
y(t) (Fi)Y*(t). In the case that y*(t) is a constant multiple can write y(t) + (Fi)siY(t). In this case we say that y(t) is multiplieativ e factor.
(5.3)
245
y(t) ~
(ri)
Y7 y(t). l
solutions
Thus for a nondegenerate path r. there are linearly independent multiplicative l (1)(t), y~2)(t) which can be combined into a solution vector solutions Yi (2) Yi (t) = {Y l)(t)' Yi (t) } such that Yi + (ri)siYi where S. is the constant diagonal matrix <s! l), s!2)>. l i i Let us now consider the case in which (5.1) has two finite singularities tl, t2 in the complex tplane and a singularity at infinity. We wish to consider (5.4)
To this end we introduce the following notation. 2Pi = s(1)z" + s(2)'i r. = s! I) s! 2), i i i M(012) = 2qi = s(1)i  s.i(2), i = 1,2,
'
where el2 C in general and is called the "link" parameter for the path FI2. Consider now a solutiQn vector YI such that YI (FI)S]YI"
(5.5)
The analytic continuation of YI about r2 yields a solution vector of (5.1) so there is a constant matrix L = (%ij) such that Y (r2)LY . The eigenvalues of L must be the path factors s~ I), s~ 2) for F2 and so
(1)
trace (L) = s2
s~2)
= 2P2
(I) s~2) = r2" det (L) = s2 This then shows that for some constant ~ the entries in L must be such that
246
%11 = P2 + ~'
%22 = P2  ~'
2 2 %12%21 = q2  ~ " As is demonstrated in [2j there is no loss in choosing = q2 cosel2 with ~ ~ Re ~12 ~ ~ and taking L to be
LI2 = p21 + q2M(Ol2 ), where I is the 2 2 identity matrix. To summarise these results we have Theorem 5.1. Let F I, r2 be nondegenerate paths. uniquely determined in the region R = {O I {0 < Re 0 < ~} u {Re 0 = 0, Im e < 0} u {Re e = ~, Im e < 0} such that there is a unique vector Yl2 determined up to a constant scalar multiplier with the properties Yl2 + (FI)SIYI2 Y12 (F2)LI2YI2 From this result we deduce Theorem 5.2. The path factors for FI2 are the roots of the equation s 2  2(plp 2 + qlq2 cos e12)s + r ir2 = O. (5.6) Then there exists a link parameter el2
Since (5.1) in this case has only 2finite singularities r12 is equivalent to  F and so the path factors for  r must be precisely the path factors
mining the link parameter el2 or the characteristic exponents of 5.1 at infinity. Consider for example Hill's equation in algebraic form, i.e.
__ 1  ~ dw ~(t) w _ 0, (5.7)
dXw + 2 {~ + dt 2
} dt + t(t
l)
247
are regular singularities while the singularity at infinity is irregular. turns out that with t I = 0, t2 = I Pl = P2 = O, ql = q2 = 1, r I = r2 = l, sll) = s~ I) = 1,
s12) = s~ 2) = I.
in the definition of the tperiodic problem (see (4.1) (4.2) above), to the descriminant D discussed in section 3.
characteristic exponents at infinity is related to the solvability of the equation D = 2 cos ~t. This interplay between characteristic exponents and the descriminant D has been exploited considerably in the study of such special functions as Whittaker functians, Mathieu functions, spheroidal wave functions etc. and in the older literature to the study of the socalled Hill determinants. Arscott Ill. See for example the book of
periodic problems is yet to be investigated. It should be noted that the assertion that there is at least one multiplicative solution of (5.7) for FI2 is simply the algebraic form of Floquets theorem. If (5.1) has three singularities in the finite part of the plane then the situation becomes much more complicated. However the results can be simply stated see [2].
The path factors for FI23 are determined as the roots of the equation s2  2s{PlP2P 3 + I qlq2P3 cos012  2iqlq2q 3 sin012 sin @23 sine3l} 1,2,3 + rlr2r 3 = O, (5.8)
248
(5.9)
In this result (5.9) has, at least for real 8ij, the following geometrical interpretation. Suppose e12, 013, e23 are the three sides of a spherical triangle, Equation (5.8) appears to lack However returning to
the symmetry one would expect between the indices I, 2, 3. our geometrical analogy we observe that N = sin~23 sin @31 sin @12
is simply the "norm" of the our spherical triangle and so is also given by N 2 = sins where 2s = 812 + 023 + 031. The syn~netry of N and so of (5.8) is now obvious. As an illustration of theorem 5.3 consider the algebraic form of Lam~'s equation, i.e. sin(s  Ol2)sin(s  023)sin(s  e31) ,
__ I I I d2w + ~ {t + dt 2 t ]
I + t ~  _ 
dw
h  ~(V + 1 ) k 2 t
w = 0.
(5.10)
dt 2
t(t
l ) ( t  k 2)
Here we take t I = 0, t~ = I, t3 = k
s (I)~. = I,
s (2)~. = I,
Pi = O,
qi = I,
r.~ = 1,
i = 1,2,3.
Thus in this case (5.8) reduces to s But FI23 is equivalent to  F expi(v + l)n. Consequently N = ~sinv~.
I
 4iNs
 I = O.
(5.11)
Thus if two of the link parameters O., are known then the third is determined by lj
(5.11).
Using ideas related to the above Arscott and Wright [3] have essentially made a study of the link parameters when ~ of the resulting solutions. is rational and have discussed the uniformi~
249 and results may have direct bearing on the outstanding problem of quasiconformal mappings introduced at the beginning of this paper.
References [l] F. M. Arscott, (1964). [23 F. M. Arscott and B. D. Sleeman, ential equations. [33 Multiplicative solutions of linear differ(1968), 263270. Periodic Differential Equations. Pergamon Press, London
[43
F. V. Atkinson, Operators.
Matrices an d Compact
Academic Press, New York and London 1972. Quasi conformal mappings with applications to differential
[5]
Lipman Bers,
equations, function theory and topology. Bull. Amer. Math. Soc. 83 (1977)
10831100.
[6]
Proceedings of Symposia in Pure Mathematics, Vol. 28 Part 2. Math. Soc. Providence R.I. (1976). [8] P. J. Browne,
problems. Proc. Roy. Soc. Edin. A (]978) (to appear). [9] P. J. Browne and B. D. Sleeman, Stability regions for multiparameter
systems of periodic secondorder ordinary differential equations (submitted), [I0] E. A. Coddington and N. Levinson, McGrawHill, New York (]955). Theory of Ordinary Differential Equations.
25O
[11]
M. S. P. Eastham,
Scottish Academic Press, Edinburgh and London, (1973). [12] A. Erd~lyi et al, New York [13] [14] [15] (1955) Dover (1956). Vol 2. 1888. Higher Transcendental Functions Vol 3. Mc GrawHill,
B. A. Hargrave and B. D. Sleeman, The numerical solution of twoparameter eigenvalue problems with an application to the problem of diffraction by a plane angular sector. J. Inst. Maths. Applics. 14 (1974) 922. 60
[16]
[17] [18]
E. L. Inee,
Dover (1956).
A. E. Pearman,
Lam~ functions in scattering problems with particular PhD Thesis, Dundee (1974).
eigenvalue problem in ordinary differential equations using variational principles. [20] J. London Math. Soc. (2) 6 (1973) 705712. Multiparameter Spectral Theory in Hilbert Space.
B. D. Sleeman,
UNIFORM SCALE FUNCTIONS AND THE ASYMPTOTIC EXPANSION OF INTEGRALS by Jet Wimp Drexel University, and University of Strathclyde, Glasgow. Philadelphia
NOTATION
S, sector : S E S(~l,~2.z o) = {z ~ ~ i~ I < arg(zz o) ~ ~2' z # Zo} , SA = S (  7 + A, ~A, 0), 0 < A < ~ .
UR(Zo) = interior of circle, radius R, centre Zo," UR(O) = UR; U 1 = U. CR(Z o) = circumference
=
CR(O)
CR;
C1
C.
~R(Zo) = UR(Zo) U
H 2 = Hardy class.
CR(Zo).
io
o
zt f(t)dt e
(Laplace transform).
252
series are today one of the next important form they were discovered (1886).
tools in mathematical
an
In their original
almost simultaneously
by Poin
community,
and which can be traced back at least to Euler series. analysis in general  is a subject that Even recently, asymptotic
use of divergent
has never quite managed to escape the taint of mystery. series have not always been spoken of clearly. to clear the dead wood of heuristic
I feel that Arthur Erd~lyi did much As Erd~lyi himself His idea of an
pointed out, his ideas were not new (what mathematical asymptotic scale goes back at least to Schmidt basis.
idea is ?).
had become the enemy of the best. onable definitions. were very reasonable the asymptotic the individual And Erd~lyi's
What were needed  among other things  were reasdefinitions of asymptotic equivalence and scale that
properties
terms, let's call them fk(z,~), where ~ is in some parameter For certain values of z individual terms could be O, thus
c ~P and z E S.
conveying a false sense of the precision of the expansion. needed were functions  call them pk(Z,~) could not become arbitrarily
 which "closely" bounded the fk but which (z,~) sets. Fortunately, the essential
small on compact
idea turned out to be easier to define and implement suggests. permissible The effect of these definitions
was to greatly enlarge our concept of a expansion. Now the fk(z,~) functions than the
were allowed to be members of a much larger class of mathematical simple inverse powers of z occurring larger the parameter expansion in~, space GL
of uniformity
to functions having the unit circle as a natural boundary  to en. series that neither Polncare i nor Stieltjes would have recognized and properties of these new base functions became subIn this paper, I wish to discuss some of the special base
the behaviour
interest
asymptotic expansions
involved.
253
2..... Poincar~ asymptotic series We differentiate between two kinds of Pozncare asymptotic series. Definition (i) Let f be defined in S.
co
By z>= in S,
(I)
we m~an
f ~
k
If 
akzk I = o(zK),
z = in S, K = 0,1,2, . . . . By in S, o
f~ w e mean
E a (ZZo)k , kO
K
z z
If 
E k=O
a {ZZo)kl. = o[(Z~Zo)
z z in S, > K = 0,1,2, . . . . .
As is well known, asymptotic series in common sectors may be added, multiplied (Cauchy product), (synthetically) divided. Any function has at most one asymptotic
expansion in a given S, but different functions may have the same asymptotic expansion. If f is analytic at Zo, its Taylor series is an asymptotic expansion. material see Knopp ~948~, or better yet, a modern treatment, such as For all this Olver C1974)o
cases of interest where f has "asymptoticlike" series for which the definition fails. A simple example is furnished by the Legendre polynomials #
= (l)n
Pn(X) 2n nl
an [(lxm)n],
dx n
n = 0,1,2,....
I will write
_,
(2) Pn(CS 8) ~ ( ) ~ k=O ( k)(
but the "~" notation can't be that of the previous definition, since individual terms are not of the required form. In fact, an even simpler example was the one that motivated Erd~lyi's first use of an asymptotic scale. Aitken, in a 1946 paper, studied some curious series. They
t The definition of all traditional special functions used in this paper will be the same as in Erd~lyi [1953].
254
(a2
I) (a2
~) 9
(3)
E k2 a2 kn+~ 
5~ (v21) (~24)
(a+k+~)
(The general
But considered
as function of the asymptotic variable n, the terms  as Aitken points out  become small, reach a minimal value, and then begin to increase again. of certain Poincar~ asymptotic expansions, see Knopp (1948). This is a feature
series could be used to accelerate the convergence of infinite series. the remainder on approximating ~2/6 by n E k=O i (k+l) 2
may be expanded in an inverse central factorial series and for large n accurately.
computed quite
In closing, Aitken says that Erd~lyi has pointed out to him (both men were at the University of Edinburgh at the time) that when a function f(t) has an expansion in t 2 k, powers of (2 sinh 7) then its Laplace transform f(z) will have an expansion (not necessarily convergent) (2k)~ P(zk) (z+k+l) in the functions I= e zt (2 sinh ~) t.2k dt. o
If z is replaced by ~, these are the functions occurring in the series (3). Gradually, in a series of papers that started with an investigation of such series,
Erd~lyi adopted the following definitions. In what follows, let ~k' ~k' fk be sequences of functions. both sequences will be defined for Izl > R, or IZZol For a given problem
Factorial series
ak/(Z+l)(z+2)...(z+k)
(1954) and his references  but not from the point of view of asymptotic
255
Definition: (i) {~k } dominates {~k } if ~k = 0(~k)' k = 0,1,2, .... (ii) {~k } weakly dominates {~k } if ~n = O(~k) for some n, (iii) ~k and ~k are equivalent if each dominates the other. k = 0,1,2 .....
k=O
fk
is an asymptotic expansion of f with respect to the scale {~k } if ~ K f  E fk = ( ~ k ) ' K = 0,1,2, .... k=O We then write
oo
(4)
f ~
E k=O
(vi) if any of the underlined words in (i)  (v) are preceded by uniformly in ~ this means the "0" or "o" signs involved held uniformly i n ~ . (vii) the series (5) f ~ E k=O Ck ~k ;
J
For the basic properties of asymptotic sequences and expansions, (1956), (1961), and particularly Erd~lyi and Wyman (1963).
see Erd~lyi
of these definitions,
an asymptotic expansion (4) loses the uniqueness property enjoy(I) or (5), see Erd~lyi (1956). A given function may This, in prac
the new expansions, we cannot expect them to do our thinking for us. ings, see Olver (1974, p.26). The reader can now make sense of the previous examples.
cussion in section 6) and the inverse central factorial series (3) is asymptotic with scale {nl2k}. 3. Choice of scale, an example This example shows how a change of asymptotic scale can make an intractable problem easy. I wish to find an asymptotic expansion for the coefficients an in
256
co
r(l+t) =
Z n=0
(I) n an t n,
Itl < i.
1 I
o
e t
(n t) n dt,
(l)n cn = ~
e t (Zn t) n dt.
Expanding gives b
t . in its = E k= 0
termwise
(l)k k~(k+l)n+l series, with scale ~k = (k+l)n since 1 ~ I e (K+2) n+l  o ~ K + I ) n]
bn
and so a n ~ Z k=O (l)k k: (k+l) n+l (1974) from a general theory. (but not to Cn~). see section 5, on the integral ; { (k+l) n } ,
an =
et
(~n t) n dt.
t* n t* = n. Everything can be carefully estimated, but when all is said and done, it is hardly
possible to give more than the first one or two terms of the expansion.
257
scales
; Laplace
initial
(Watson's
Let ~ > 0, R e B
f ~
Z k=0
z~
The applications For example, if I(z) = [ ezh(t) ~F where g,h c ~ ( B ) , g(t)dt , of Watson's lemma are many. For a discussion, see Olver (1974).
of steepest
descents
supposes
by the values
point,
eral idea is to make the substitution h(t)  h(t*) = (tt*) 2 h''(t*) 2 + ... = w 2 ,
< O. so in a neighbourhood of 0
invertible,
t = t* + [2/h''(t*)] w + .... One would expect the major contribution to I to occur at w = O. Thus
ezw2
u(w)dw,
kI 2,
Z > oo
in SA ,
258
(1947),
Erdelyl
introduces theorems
including
speaking, sequence as
z ~ in SA and vice versa. then {~n } is an asymptotic paper, Erd~lyi states paper.
sequence
And if ($ k } is an asymptotic sequence as t ~ in R and vice versa. showing when this is true, of course,
2 theorems
and correcting
induces
a correspondence
expansions
We quote two of the main results: Theorem: (generalized initial value and 0 +. theorem)
(6)
f ~
E fk ; {t k=O
} , t
; {z
~k
} ,
z ~ in some SA.
(generalized
theorem)
Z fk ; {t k=O
t ~
in R +.
Z ~k ; {z k=O
~k}
, z 0 in some SA.
each SA is a bother.
But Erd~lyi
gives an alternative
k} for which
(io)
E Ck(let)k k=O
; {tk},t 0 +
{z k}
z ~ ~ in some SA
following
equation
(2))
259
(II)
f~
E k=O E k=O
ck(etl)k
{tk},
t + 0+
f~
k:Ck/Z(zl)...(zk);
{z k1 } z +
in some SA
f ~ (12) ~
I k=O E k=O
O+ ; 2kl~ ~, z ~
(2k) l Ck/(Zk)(zk+l)...(z+k)
; {z
in some S A.
Next, Erd~lyi gives a general theorem, similar to (6)  (7), for asymptotic expansions with respect to the scales {(in t) Bk t ~kl} and {(in z) ~k z ~k} . orem generalizes a number of results given in Doetsch (19501956). Watson's lermna. The books by This the
Many authors have discussed other generalizatio~of Olver (1974), Bleistein and Handelsman above provide much information.
(ta) % g(t)dt,
x~ , y~ ,
are treated, and also a much longer survey article (1955,56) by the same author. also vander Waerden (1951), and Wong and Wyman (1972).
See
For a discussion of the numerical error involved in using Poincare type asymptotic series, Olver's book (1974) is excellent. 5. Darboux's method Let f e ~(0). (13) f = It is no loss of generality here to assume f e ~(U) at least, so See also the recent paper by Pittnauer (1973).
usually handled on an ad hoc basis by applying the method of steepest descents, or one of its variants, (14) fn
=
to the integral I ~ [ f(t) ~ r t dt, Such an approach does not usually yield a complete
where F c U is homotopic to C.
expansion, and the details may be very messy. The kind of argument used is wellillustrated p. 329) where f(t) = exp[et]. f = Pe Q, P,Q polynomials. in an example given by Olver (1974,
The case f = eQ, Q a polynomial, was more fully treated by to earlier work. (1968). Other examples Often the asymptotic
have been given by Rubin (1967) and Harris and Schoenfeld formulasobtained
260
equations involving n and seldom is it possible to do more than derive a leading term for f n On the other hand, when f has singularities on the circle of convergence and a function g can be found which matches the behaviour of f at these points and whose Taylor's series coefficients gn are known, then a very elegant method due to Darboux In practice, what / results is often a complete asymptotic description, but not one of Poincare type, for f . Since Darboux's method has not received full attention in any of the available n texts on asymptotics and is a rich source of general asymptotic expansions, I will discuss it in some detail. Let f c ~ ( U ) and put (1878) provides an asymptotic estimate of fn in terms of gn"
f( 1 reiO) 12d0}'2,
0 < r < I.
Example: Let f = h(~t) O, ~ e C, then f g H 2. Definition: Let f, g g ~(U) and for a fixed m = 0,1,2,..., f(m)  g(m) E H 2 . a comparison function of order m (to f). In what follows let g = E n=O gn tn" Then g is called Re o > I, h E ~(U);
Theorem:
(Darboux's method)
Let g be a comparison function of order m to f. Then (15) Proof: fn = gn + (nm)' n ~. I may write I 2~i(n_m+l) m I CR h(m)(t) dt tn+l.m ,
(16)
fng n
h = fg, O<R<I,
261
But the radial limit of h (m) exists almost everywhere and eL2(C), see Rudin (1966, p.366). Thus the integral on the right of(16) maybe expressed as an integral around C
(again, see Rudin). The use of the RiemannLebesgue l e n a then gives the theorem. A simple but important case is when the singularities of f on C are finite and algebraic in nature, so (17) f
=
Z k=O
ak(r) (I
t) r
Br+kY r
' ~ ~ C,
r = 1,2,...,R;
Re m r > O,
Then the function formed by adding together the first (K+I) terms of each of the R series (17) will be a comparison function of order m provided (18) m < + min {(K+I) Rear + Re8 r}
r
I have thus demonstrated Szeg~'s result (1959, p.205) it. Theorem: (Darboux's method for algebraic singularities) Let f be as in (17). Then (19) fn = K R Z k=O r=l ~r)~r+~Tr)(_~r)n + o(nm)
for all m satisfying (18). It is easy to show this formula is uniform for ~ = (el,~2 .... ,=k) e ~[whenever ~ i s a Cartesian product of compact disjoint subsets of ~. The general term of the sum in (19) is [( r)] k~rBrI Br + ~Y = O(n ), 0 n but we have no way of interpreting the formula in terms of Pozncare s definition of an asymptotic series. However when K ~ ~ the series yields an asymptotic expansion in Pk }~Pk = l+min{k ReYr+ReBr}. r (The reader will verify this is, in fact, an asymptotic scale, since Rey r > O).
JV
It provides asymptotic
expansions for the Jacobi polynomials (SzegS, (1959)), the generalized Bernoulli polynomials, and even for more exotic polynomials, such as the Pollaczek polynomials,
For this expansion, we make the branch cut along B =[~r' ~ r ] " t s Ud(~r), t ~ B.
262
P% (x; a,b), Szeg~ (1959, p.390). Darboux himself (1878) originally applied his n method to the Legendre polynomials. Darboux's method says if we can solve the problem of finding an asymptotic formula for the Taylor coefficients of a comparison function, we can find, to within algebraic for f. This sometimes suffices to describe f n
An enormous amount of work has been done on deriving asymptotic formula for gn for certain typical g. C is exp {%/(lt)}. One of the simplest functions having an essential singularity on
Perron (1914), in fact, considered the function (lt)aexp{%/(lt)} Wright (1932) gave the complete asymptotic develop
Perron generalized his own work in (1920) to find that if (lt) a ~Qa;c l%t) = n ~0 gnt ' n then gn = ~ F(c) 1 X[I + O(n ")j. Faber (1922), Ha~sler (1930), Wright (1933, 1949) continued this kind of research, I %4 c 2 % 2 % > O,
c 3 a  ~  ~ %~n 2 e
the last reference giving the leading term of gn for the very general g(t) = (lt)a[n(lt)] b eP(t) h(t), where h ~ ~(U),
M
h(1) # O, and
c
m
e(t) = m=l
(lt) dm
Recently, Wong and Wyman (1974) have done work on functions with logarithmic singu ~ larities on the circle of convergence. Suprisingly, mathematicians have obtained results even in cases where f has singu
larities which are dense on C, i.e. when the unit circle is a natural boundary for f, Such cases are of great interest to number theoreticians. (1937), Ingham (1941), Szekeres have contributed work, Uspensky (1920), Rademacher
(1953), and Bender (1974) are some of the workers who In particular, I want to talk about Rademacher's
to this research.
triumphs of analytic number theory and a tour de force of complex analysis. Let Pn be the number of ways n can be written as a sum of positive integers $ n, Po = I, Pl = I. and Then the infinite product I write below converges for It] < 1
263
f(t) =
H (ltJ) I = E pn t , j =i n=O
Itl < i.
see Rademacher
(1973).
extends
(14) over a doubly indexed sequence of socalled Farey arcs adjoining a circle which approaches the unit circle from the inside. His analysis involves the transformation but
theory of modular functions and a lot of bone breaking estimates of integrals, the result is worth it. (20) Pn = ~ I He finds k d I sinh )',_ C = , N = (n  ~ ~i
kZl A~,n
2~in/k
Ak'n
where ~h,k is a 24k th root of unity. The series (20) is asymptotic in n (perhaps the reader can verify a scale is N2e cN/k) and also convergent, can thus be used to compute Pn" a very unusual feature in series derived this way, It
P721 and shows they agree with earlier estimates baked on the work of Hardy and Ramanuj an. What about uniform asymptotic expansions? ~ ~C ~P, In other words, if f E f(t,~),
can Darboux's method be made to yield expansions which are uniform for A typical question is, what happens if
= C
in (19)?
As the reader may suspect, such questions are very difficult to answer, and
the requirement of uniformity of the expansion in a parameter set invariably raises the hierarchy of the base functions in the expansion for f . See Olver's very nice n (1975) discussion of this "youcan'tgetsomethingfornothing" principle of asymptotic analysis. The only effort I know in this area is the substantial work of Fields ~968),
who treats the case when f in (19) has two singularities which are allowed to coalesce. Finally, can anything at all be said when f belongs to some general class of funcSurprisingly,
tions more interesting than just those with algebraic singularities? yes. Hayman (1956), Wyman (1959), and Harris and Schoenfeld
the problem of defining what general properties of classes of functions ~ , adequate to enable one to make asymptotic statements about fn for f g ~ . call the elements of ~ admissible functions.
not easy to describe, but generally they involve restrictions of the elements.
264
6. Higher transcendental
scale s
Let us return to the problem of estimating (21) I(z) = [ ezh(t'~) g(t)dt, ~p ~ e~C~ P.
It would seem that if I could do the analysis for certain simple representative
choices those
of h then I have really solved the problem for a wide class of integrals, namely, which can be reduced to the representative
method of steepest descents enabled me to reduce an integral with one stationary critical point to an integral which could be handled by Watson's lemma (h(t) = t). The simplest function having a movable critical point is h(t,~) = st + t2, and the representative I(z) = integral is g(t)dt.
=o e_Z (~t+t2)
O
g=
E gk tk, k=O
Itl <~,
(22)
E k=O fk =
gk fk'
(23)
~ e_ z (at+t 2)
O
tk dt.
St seems fk cannot itself be uniformly estimated in ~Lby simpler functions. this is to be expected: as the requirements
But
the dimensionality of the parameter space ~ p increase, so will the complexity of the base functions involved in the asymptotic expansion. Nevertheless, the functions fk can be considered known. They can be expressed in
term recurrence relation, and they can be very easily generated on a computer by applying the Miller algorithm,(see Wimp (1970) and the references given there.) have 2zfk+ 2 + z~fk+ 1  (k+l)f k = O, k = O,1,2, ..... We
265
aN
~I + O(k)]
k ~,
and using this I can show if ~ E (0,~) the Miller algorithm for the computation of fk will converge for z e SAOne would expect the expansion (22) to be of Polncare type since Theorem:
+
J
fk is a uniform
asymptotic
scale
in a as z ~ in R .
I [//2 t* = ~ 4 I get
8k +z
~]
Ik = ~k I
l+u) e
du,
~k = (t*)k+l e
~zt*zt .2
0~<o,<
i.
Thus the integral above may be bounded and bounded away from zero uniformly in ~ and z, so it follows that fk+l fk 7 '
z ~ ~ > O,
independent of a and z. Rather than show for which conditions the expansion C22) is a Polncare expansion, I will use an asymptotic scale simpler than fk (but still uniform i n k ) . result is in Erdeflyi (1970). Theorem: Let g e ~(0) and let I exist for some z > 0 and all ~ e ~ then I ~ ~ ~g~ fk ; {(~z + 2/~z)k} , z + ~ in R +. kO The following
J
266
i
I have taken T = ~ in Erdelyl's result, and also assumed g independent of z. Note that the absolute convergence of the Lebesgue integral guarantees that his hypothesis (d) is satisfied, as an integration by parts of (zz o)(~t+t 2) e
e
it G(t)dt, G(t) =
o
z (~u+u2) e o
g(u) du
will show. For additional material on other such expansions, see Erd~lyi (1974). The next level of difficulty is encountered when h in the integral (21) has two movable critical points, the dimension of the parameter s p a c e ~ still being I: (24) h(t,~) = a(~)t + b(~)t 2 + t3, e ~ ~.
Under suitable conditions l(z) may be expressed as a sum of two asymptotic series with scales 2 Ai(cz ~) 2k
Z
2 Ai'(cz ~) 2k
Z
'
respectively, where c depends on ~ and Ai, Ai' are Airy functions. They may be i expressed in terms of modified Bessel functions of the second kind, order ~ and 2 order 7' respectively, see Olver (1974, p.392 ff.) An analysis of integrals which can be reduced to this form by a change of variable constitutes the famous method of Chester, Friedman and Ursell (1957), (CFU). For an
exposition of this method, see the survey by Jones (1972) orOlver (1974).
Olver~ in
a series of papers that are now considered classics (1954a, 1954b, 1956, 1958) encountered their same functions in determining asymptotic expansions for the solutions of sound order linear differential equations with large parameter in the neighbourhood of a turning point. For those functions to which it applies, Olver's The CFU
theory has the advantage that z may approach ~ in sectors SA other than R.
theory establishes a nice relationship between the asymptotic expansion of integrals and the asymptotic expansion of the solutions of differential equations. Determining the precise sregion of uniformity of the CFU expansions, and finding conditions guaranteeing that an integral may be transformed into one which can be handled by the CFU technique are very difficult problems, and Ursell devoted two subsequent papers to these investigations (1965, 1970). At least the base functions
in the expansion, the Airy functions (25) are well understood and can be easily calculated on modern computers. If one wants to analyze the integral l(z) = IpeZH(w'~ ) G(w,~)dw
(25)
267
where H is to be transformed into the general polynomial h(t,~) = ~i t + ~2 t2 + ... + ~ tp + tp+I, p then one will have to live with incomplete results; (25) to the representative integral f= ~ = (~i,~2 .... ,~p ),
Jo eZh(t'~)
g(t,~)dt
involves difficulttoverify hypotheses, and some of the work is only formal. who have treated this problem are Bleistein (1966, 1967) and Ursell (1972). case the base functions are called generalized Airy functions.
Authors In this
They satisfy a
differential equation of order p+l (see Bleisten (1967)), possess an asymptotic expansion in z (Levey and Felsen (1969)) and the techniques Wimp uses on similar integrals (1969) will work to show the functions satisfy a(p+2) term recursion relationship to which the Miller algorithm can be applied to compute the functions. The real
problem, though, is not the analyzing the properties of the base functions, but justifying the transformation of the given integral to representative form. Obviously, precise information about the asymptotic expansion of the very general integral l(z) = IrH(z,t,g)dt
For integrals such as these, there are a large number of Often it is assumed that z is real, and F = ~ 0 , ~ , and Over the last decade an
enormous number of relevant articles by E. Riekstins and other authors have appeared in the somewhat obscure publication Latvian Mathematical Yearbook. book See also the
(1974) by Riekstins, the book by Bleistein and Handelsman (1975) and papers by It is my
the authors Handelsman, Lew and Bleistein (1969, 1971, 1972, 1973).
personal feeling that a unified treatment of such integrals will involve a large number of complex and all but unverifiable hypotheses on the function H. Perhaps
the whole of asymptotic analysis of integrals (the same could be said of differential equations and difference equations) has reached the point of diminishing returns. The physicist waves his hands and obtains an asymptotic expression which he uses with confidence because he "knows" it must be ture. For difficult problems the mathematPerhaps for those problems 
say, integrals with coalescing multiple critical points and singularities  we are couching the answer in the wrong terms, and it is tempting to hope that there might exist a choice of base functions  such as in the example in section make the impossible easy. 3  that would
268
REFE&ENCES Aitken, A.C. (1946). A note on inverse central factorial series, Proc. Edinburgh Math. Soc. ~, 168170. Bender, E.A. (1974). Asymptotic methods in enumeration. SIAM Rev. 16, 485515. Bleistein, N. (1966). Uniform asymptotic expansions of integrals with stationary point and nearby algebraic singularity. Cormm. Pure Appl. .Math.. . . 353370. . . . . 19, Bleistein, N. (1967). Uniform asymptotic expansions of integrals with many nearby stationary points and algebraic singularities. J. Math. Mech. 17, 533560. Bleistein, N., and Handelsman, R.A. (1972). A generalization of the method of steepest descent. J. Inst. Math. Appl. iO, 211230. Bleistein, N., and Handelsman, R.A. (1975). Asymptotic expansions of integrals. Holt, Rinehart, Winston, New York. Chester, C., Friedman, B., and Ursell, F. (1957). An extension of the method of steepest descent. Proc. Cambridge Philos. Soc. 53, 599611. Corput, J.G. van der (1934). Zur Methode der stationaren Phase, I. Comp. Math. i, 1538. Corput, J.G. van der (1948). On the method of critical points, I. Nederl. Akad. Wetensch. Proc. Ser. A. 51, 650658. Corput, J.G. van der (1955/56). Asymptotic developments, I. Fundamental theorems of asymptotics. J. Analyse Mat h. ~, 341418. Darboux, G. (1878). Memoire sur l'approximation des fonctions de tr~sgrands nombres. J. Math. Pures App !. (3) ~, 556, 377416.
I!
Doetsch, G. (195056).
Erdelyl, A. (1947). Asymptotic representation of Laplace transforms with an application to inverse factorial series. Proc. Edinbursh Math. Soc. 8, 2024. Erdelyl, A. (1956).
I .
Erd~lyi, A. (1961). General asymptotic expansions of Laplace integrals. Arch. Rational Mech. Anal. ~, 120. Erdelyl, A. (1970). Uniform asymptotic expansions of integrals, in Analytic methods in'mathematical physics (R.P. Gilbert and R.G. Newton, eds) Gordon and Breach, New York, 149168. Erd~lyi, A. (1974). Asymptotic evaluation of integrals involving a fractional derivative. SlAM J. Math. Anal. ~, 159171. Erd~lyi, A., et. al. (1953). York.
J .
Erdelyl, A. and Wyman, M. (1963). The asymptotic evaluation of certain integrals. Arch. Rational Mech. Anal. 14, 217260.
269
Faber, G. (1922). Abschatzung von Funktionen grosser zahlen, Bayerischen Akad." Wiss. Munchen Math. phys. Klasse, 285304. Fields, J.L. (1968). A uniform treatment of Darboux's method. Arch. Rational Mech. Anal. 27, 289305. Handelsman, R.A., and Bleistein, N. (1973). Asymptotic expansions of integral transforms with oscillatory kernels : a generalization of the method of stationary phase. SlAM J. Math. Anal. 4, 519535. Handelsman, R.A., and Lew, J.S. (1969). Asymptotic expansion of a class of integral transforms via Mellin transforms. Arch Rational Mech. Anal. 35, 382396. Handelsman, R.A., and Lew, J.S. (1971). Asymptotic expansions of a class of integral transforms with algebraically dominated kernels. J. Math. Anal. Appl. 35, 405433. Hardy, G.H. and Ramanujan, S. (1918). Asymptotic formulae in combinatory analysis. Proc. London Math. Soc, (2) 17, 75115. Harris, B., and 8choenfeld, J. (1968). Asymptotic expansions for the coefficients of analytic functions, illinois J. Math. 12, 264277.
t~ TI
Hausler, L. (1930). Uber das asymptotische Verhalten der Taylorkoeffizienten einer gewissen Funktionenklasse Math. Zeit. 32, 115146. Hayman, W.K. (1956). A generalization of Stirling's formula. J. Reine. Ang. Math. 196, 6795. (1941). A Tauberian theorem for partitions. Ann. Math. 42, 10751090.
Ingham, A.E.
Jones, D.S. (1972). Asymptotic behaviour of integrals. SIAM Rev. 14, 286316. Knopp, K. (1948). Theory and application of infinite series. Hafner, New York.
Levey, L., and Felsen, L.B. (1969). An incomplete Airy function and their application to diffraction problems. Radio Science 4, 959969. Moser, L., and Wyman, M. (1956). Asymptotic expansions, I. Canad. J. Math. 8, 225233. Moser, L., and Wyman, M. (1957). Asymptotic expansions, II. Canad. J. Math. 9, 194209. Norlund, N.E. (1954). Vorlesungen uber Differenzenrechnung. Chelsea, New York.
Olver, F.W.Jo (1954a). The asymptotic solution of linear differential equations of the second order for large values of a parameter. Phil. Trans. Roy. 8oc. London. (A) 247, 307327. Olver, F.W.J. (1954b). The asymptotic expansion of Bessel functions of large order. Phil. Trans. Roy. Soc. London. (A) 247, 328368. Olver, F.W.J. (1956). The asymptotic solution of linear differential equations of the second order in a domain containing one transition point. Phil. Trans. Roy. Soc. London. (A) 249, 6597.
270
Olver, F.W.J. (1958). Uniform asymptotic expansions of solutions of linear second order differential equations for large values of a parameter. Phil. Trans. Roy. Soc. London. (A) 250, 479517 Olver, F.W.J. (1974). Asymptotics and special functions. Academic, New York.
Olver, F.W.J. (1975). Unsolved problems in the asymptotic estimation of special functions, in Theory and application of special functions (R. Askey, ed.) Academic, New York, 99142.
11
Perron, O. (1914). Uber das infinitare Verhalten der Koeffizienten einer gewissen Potenzrei[~e. Arch. Math. Phxs. 22, 329340.
it
Perron, O. (1920). Uber das verhalten einer ausgearteten hypergeometrischen Reihe bei unbegrenztem Wachstum eines Parameters. J. fur Math. 151, 6378. Pittnauer, Franz (1973). Zur Fehlertheorie positiver asymptotische Potenzreihen. Math. Nachr. 58, 5561. Polncare, H. (1886). Sur les integrales irregulleres des equatlons lineares. Acta Math. 8, 295344. Pollya, G. (1922). Uber die Nullstellen sukzessiver Derivierten. Math. Z. 12, 3660. Rademacher, H. (1937). On the partition function p(n). Proc. London Math. Soc. (2), 43 241254. Rademacher, H. (1973). Topics in analytic number theory. SpringerVerlag, Berlin.
Riekstins, E. (1974). Asymptotic expansions of certain types of integrals involving logarithms. Latvian Math. Yearbook 15, 113130. Riekstins, E. (1974). Asymptotic expansions of integra! s, v. I. (Russian) Riga. Latvian Math. Yearbook.
Rubin, R.J. (1967). Random walk with an excluded origin. J. Mathematical Phys. 8, 576581. Rudin, W. (1966). Real and complex analysis. McGrawHill, New York.
Schmidt, H. (1936). Beitrage zu einer Theorie derallgemeinenasymptotische Darstellungen. Math. Ann. 113, 629656. Stieltjes, T.J. (1886). Recherches sur quelques ser~es semiconvergentes. Ann. Sci. Ecole. Norm. Sup. (3) ~, 201258. Szego, G. (1959). Orthogonal polynomials. Amer. Math. Soc.,New York.
I
Szekeres, G. (1953). Some asymptotic formulae in the theory of partitions. II Quart. J. Math. (2) ~, 97111. Ursell, F. (1965). Integrals with a large parameter. The continuation of uniformly asymptotic expansions. Proc. Cambridge Phil. Soc. 61, 113128. Ursell, F. (1970). Integrals with a large parameter : paths of descent and conformal mapping. Proc. Cambridge Phil. Soc. 67, 371381. Ursell, F. (1972). Integrals with a large parameter. Several nearly coincident saddle points. Proc. Cambridge Phil. Soc. 72, 4965.
271
Uspensky, J.V. (1920). Asymptotic expressions of numerical functions occurring in problems of partitions of numbers. Bull Acad. Sci. Russie (6) 14 199218. Waerden, B.L. van der (1951). On the method of saddle points. Appl. Sci. Res. (B) 2, 3345. Wimp, J. (1969) o On reeursive computation. WrightPatterson AFB, Ohio. Rep. ARL69  O186, Aerospace Res. Lab.
Wimp. J. (1970). Recent developments in recursive computation, in SIAM studies in applied mathematics , v.6,. SIAM Philadelphia, 110123. Wong, R., and Wyman, M. (1972). A generalization of Watson's lemma. Canad. J. Math. 24, 185208. Wong, R. and Wyman, M. (1974). Wright, E.M. The method of Darboux. J. Approx. Theory IO, 159171.
(1932). The coefficients of a certain power series. J. London Math. Soc. ~, 256262. (1933). On the coefficients of a power series having exponential singularities, I. J. London Math. Soc. 8, 7179. (1949). On the coefficients of power series having exponential singularities. II. J. London Math. Soc. 24, 304309.
Wright, E.M.
Wright, E.M.
Wyman, M. (1959). The asymptotic behaviour of the Laurent coefficients. Canad. J. Math. II, 534555.