H. Stichtenoth M. A. Tsfasman (eas) Coding Theory and Algebraic Geometry Proceedings of the International Workshop held in Luminy, France, June 17-21, 1991 Springer-Verlag Betlin Heidelberg. New York London Paris Tokyo Hong Kong Barcelona BudapestCQ VEN eV a we od er Henning Stchtenoth Neowe ge le cod. S Fachbercich 6 Mathematik und Informatik Coby r= 8605" Universitit GHS Essen Universiitsste 3, W-4300 Essen 1, Fed. Rep. of Germany Michael A. Tsfasman Institute of Information Transmission (IPP, 19, Ermolovoi st, Moscow, GSP ~ 4, 101447, Russia groan oigclov co. ‘Mathematics Subject Classification (1991): 14-06, 94-06, 11-06 ISBN 3-540-55651-6 Springer-Verlag Berlin Heidelberg New York ISBN 0-387-55651-6 Springer-Verlag New York Berlin Heidelberg ‘This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting reproduction on microfilms or in any other way, ‘and storage in data banks. Duplication of this publication or pats thereof is permitted ‘only under the provisions of the German Copyright Law of September 9, 1965, in its ‘current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the German Copyright Law. © Springer-Verlag Berlin Heidelberg 1992 Printed in Germany ‘Typesetting: Camera ready by author/editor Printing and binding: Druckhaus Beltz, Hemsbach/Bergst. 46(3140-543210 - Printed on acid-free paper Foreword ‘The workshop "Algebraic Geometry and Coding Theory - 3” or by the Insti- tute of Information ‘Transmission (Moscow), University of Essen, Equipe Arithmeétique et Théorie de IInformation de C.N.R-S. (Marseille-Luminy), and Group d’Btude du Codage de Toulon took place in the Centre International de Rencontres Mathematiques, Sune 17-21, 1991, ‘The workshop was a continuation of AGCT-1 and AGCT-2 that took place in 1987 and 1989, respectively. It is to be followed by AGCT-4 in 1993, etc., each time held in C.LRM. ‘The list of participants follows. tis our pleasure to thank the staff of C.1.R.M. for their hospitality, the participants for their interest, all supporting organizations for their financial support, and Springer- Verlag for the Proceedings. Organizers, HStichtenoth M.Tsfasman G.Lachaud J.WolfmannAGCT 8 - List of Participants Aubry, Yves (Marseille) Blahut, Richard E. (Owego, N-Y.) Boutot, Jean-Prangois (Strasbourg) Bruen, Aiden (London, Ontario) Carral, Miche! (Toulouse) Chassé, Guy (Issy les Moulineaux) Cherdieu, Jean-Pierre (Guadeloupe) ‘Cougnard, Jean (Besangon) Deschamps, Mireille (Paris) Driencourt, Yves (Marseille) Duursma, Iwan M. (Eindhoven) Ehrhard, Dirk (Diisseldorf) Gillot, Valérie (Toulon) Guillot, Ph. (Genevilliers) Hansea, Johan P. (Aarh Harari, Sami (Toulon) Hassner, Martin (San Jose, Ca.) Helleseth, ‘Tor (Bergen) Heholdt, Tom (Lyngby) Katsman, Gregory (Moscou) Kumar, Vijay (Los Angeles) Kunyavskii, Boris E. (Saratov) Lachaud, Gilles (Marseille) Langevin, Philippe (Toulon) Lax, Robert (Baton Rouge) Le Brigand, Dominique (Paris) Li, Winnie (Pennsylvania State) Lopez, Bartolomé (Madrid) Luengo, Ignacio (Madrid) Michon, Jean-Francis (Paris) Munuera, Carlos (Valladolid) Nogin, Dimitri (Moscou) Pedersen, Jens Peter (Lyngby) Pellikaan, Ruud (Eindhoven) Perret, Marc (Marseille) Polemi, Despina (New York) Rodier, Francois (Paris) Rodriguez, M.-C. (Madrid) Rolland, Robert (Marseille) Rotillon, Denis (Toulouse) ‘Seguin, Gerald (Kingston, Ontario) Serre, Jean Pierre (Paris) Shahrouz, Henri (Cambridge, Ma.) Shparlinski, Igor (Moscou) Skorobogatov, Alexei (Moscou) Smadja, René (Marseille) Sole, Patrick (Velbonne) Stichtenoth, Henning (Essen) Stokes, Philip (Valbonne) ‘Thiongly, Augustin (Toulouse) ‘Tsfasman, Mikhail (Moscou) Viadut, Serge (Moscou) ‘Voss, Conny (Essen) ‘Wolfman, Jacques (Toulon) Contents Hi Stichtenoth, M.A. Tsfasman: Algebraic Geometry and Coding Theory. An Introduction Y. Aubi Proje Reed-Muller Codes Associated to Algebraic Varieties D. Erhard: Decoding Algebraic-Geometric Codes by Solving a Key Equation G. Frey, M. Perret, H. Stichtenoth: On the Different of Abelian Extensions of Global Fields A. Garcia, R, Lax: Goppa Codes and Weierstrass Gaps N. Hamada, T. Helleseth: On a Characterization of Some Minihypers in PG(t,q) (q= 3 or 4) and its Applications to Error-Correcting Codes J.P. Hans Group Codes Deligne-Lusztig Varieties and G.L. Katsman, M.A. Tsfasman, $.G. Vladut: Spectra of Linear Codes and Error Probability of Decoding P.V. Kumar, K.Yang: On the True Minimum Distance of Hermitian Codes BLE. Kunyavskii: Sphere Packings Centered at S-units of Algebraic ‘Tori J.P. Pedersen: A Function Field Related to the Ree Group 18 26 33, 108 122vu R, Pellikaan: On the Gonality of Curves, Abundant Codes and Decoding LE, Shparlinksi, M.A. Tsfasman, $.G. Vladut: Curves ‘with Many Points and Multiplication in Finite Fields P. Stokes: The Domain of Covering Codes M.A. Tsfasman: Some Remarks on the Asymptotic Number of Points C. Voss: On the Weights of Trace Codes F. Rodier: Minoration de Certaines Sommes ‘Exponentielles Binaires A.N. Skorobogatov: Linear Codes, Strata of Grassmannians, and the Problem of Segre 132 a5 170 178 193 19 20 Algebraic Geometry and Coding Theory An Introduction Henning Stichtenoth, Michael A. Tsfasman H.St.: Fachbereich 6 - Mathematik, Univ.GHS Essen, D-4300 Essen 1, Germany M.To.: Institute of Information ‘Transmission, 19 Ermolovoi st., Moscow GSP-4, U.S.S.R. About ten years ago V.D.Goppa discovered an amazing connection between the theory of algebraic curves over a finite field F, and the theory of error-correcting block qrary codes. The idea is quite simple and generalizes the well known construction of Reed-Solomon codes. ‘The latter use polynomials in one variable over F and Goppa generalized this idea using rational functions on an algebraic curve. Here is the definition of an algebraic geometric code (or a geometric Goppe code). Let X be an absolutely irreducible smooth projective algebraic curve of genus g over F,. Consider an (ordered) set P = {Pj,...,Pq} of distinct F,-rational points on X and an F,-divisor D on X. For simplicity let us assume that the support of D is disjoint from P. The linear space L(D) of rational functions on X associated to D yields the linear evaluation map Evp :L(D) > FP Fr (F(Pi)s--+s flPa)) ‘The image of this map is the linear code C = (X,P,D)1, we study. ‘The parameters of such a code can be easily estimated. Indeed, let P = Pi..+-Pa, then the dimension k is given by #=(D)-4D-P) and in particular if 0 < deg D n+1—g which is by g worse than the simplest upper bound valid for any codektdSndl ‘An equivalent description of these codes can be given in terms of algebraic function fields in one variable over Fy. The curve X corresponds to the function field F = Fy(X), and F,-points on X correspond to places of F of degree one. Originally, Goppa used the dual construction using differentials on X rather than functions, and the residue map. Unfortunately, there are at least two different traditions of notation. The second ‘one uses D for our P and G for our D , and the code is denoted Cx(D,G). ‘The construction can be generalized in several directions. In particular one can use sheaves (or some other tricks) to avoid the condition PASuppD = 9. The generalization to the case of higher dimensional algebraic varieties looks very promising but so far the results are few. ‘There are several main streams of the development of the theory. Let us briefly discuss some of them. Asymptotic problems. One of the fundamental problems of coding theory is to construct long codes with good parameters (rate and relative minimum distance). One of the starting points of the theory was the construction of long codes which are asymptotically better than the Gilbert-Varshamoy bound. The other asymptotic question is which codes can be constructed in polynomial time. Specific curves. There are many interesting examples of curves with many Fy points which lead to codes with good parameters. Sometimes such curves and codes have nice additional properties, such as large automorphism groups. Spectra and duality. The study of weight distribution and of duality leads to interesting questions of algebraic geometry, such as the study of Weierstrass points and special divisors on a curve. Decoding. Surprisingly enough the decoding problem can be set in purely alge- braic geometric terms and again one needs information about special divisors. Exponential sums. Another component of the picture is the theory of exponential sums closely related both to algebraic geometry and to coding theory. Related areas. ‘The theory of algebraic geometric codes has either analogues ‘or applications in several other topics. Such are sphere packings and spherical codes, multiplication complexity in finite fields, graph theory, and s0 on. ‘These applications also require eubtle information about the goomotry and arithmetic both of function fields and of number fields, ‘To conclude, the first ten years of development show that the connection between algebraic geometry and coding theory proves fruitful for both, giving new results and posing many exciting questions. __ Several books and many papers on the subject are either published or in prepa- Tg, Tte paar re too numerous to Tit them Bere and we refer othe extenive in [T9/VI] and to references given in the papers of this v. is the list of books. a popes omme. Here [Ge] V-D.Goppa, Geometry and Codes. Kiuwer Acad. Publ., 1988 [Mo] C.J. Moreno, Curves over Finite Fields. Cambridge Univ. Press, 1901 [Sti] H.Stichtenoth, Algebraic Function Fields and Codes. Springer-Verlag (in preparation) [Zs/Vi]_ M.A.Tsfasman, S.G.Vladut, Algebraic Geometric Codes. Kluwer Acad. Publ., 1991 [vG/vE] J.H.van Lint, G.van der Geer, Linear Codes and Algebraic Curves, Birkhauser, 1988Reed-Muller Codes Associated to Projective Algebraic Varieties ‘Yves AUBRY Equipe CNRS “Arithmétique et Théorie de information” CLM. Laminy Case 916 - 13288 Marseille Cedex 9 - France, Abstract ‘The clasical generalized Rood Mle codes inwodued by Kasai Lin and Peterson (5), and sui also by Dele, Goethals and Mac Willams (2), ae defined over the afin space A'(P) over the finite eld Fg ‘ith q elements. MoroverLachaud (6), following Mani and Vlad (7, has considered projective Reed-Muller, coves, ie defined over the projective space PVF). In his paper, the evaluation ofthe farms with coefficient in the Fite eld Ris made on the poins of a projective algebraic variety V over the projective space P(F,). Firstly, we consider the case where V is 1, then Qc H ean be seen 45 a quadric in a space of dimension n 1. We know (see for example (8)) that the ank of Qo Hist ~1orr—2, Then, either Q7 His nondegenerate (in P®-1(F,), or Qn His of rink t—2=n— 1 (whence degenerate in P®~1(Fy) ; one says in this last case that His zangent 0Q[Now we have to know what isthe value of «(1 H), ic. what happens to the type of te quadric. Ifthe hypesplane His not ngent to Q, its obvious that Q 7H becomes parabolic if Q is hyperbolic or elliptic (indeed r(Q) is necessarily even, and if H is not tangent we have QAH) =1(Q)~ I hence odd, then Q.- His parabolic); and Q/> H becomes hyperbolic or clliptic if Qs parabolic (same reason rest onthe parity ofthe ranks). ‘Now ifthe hyperplane H is tangent to Q, we have the following proposition (se [13)) = Proposition 2 : The quadric Q 7H is of the same type as the nondegenerate quadric Q if the hyperplane His tangent to Q. “Then, we can give the result about the hyperplane sections of a quadric of any rank : ‘Theorem 2 : Let Q be a quadric of P"\(F,) of rankr whose decomposition is, Qa Var V Ort and let H be a hyperplane of P®(F,). Then: a) IFH> Vy_, then 1QAH1= Aq_2 4 (W1Q-1 0H)= D g8-F- D7 fH, is not tangent to Q,_,, and IQAH1=mq-2 + (HQ= Dg? if Hy, is tangent 10 Q,_, where H,. is the hyperplane of P*~"(Fq) defined by Hy = Zpy_ 6) where h is the linear form in Fq(Xq,..X-_1}y defining H; moreover o(Q,_ 1 0 H,) is equal to 1 if Qis hyperbolic or elliptic, and equal to 0 or 2 if Qis parabolic. bP Var then VQ = Hq_2 + (CQ) — 1) g™9-F-?, We suppose that the r variables appearing in the quadratic form F defining Q are Xpats If we set Hi; the hyperplane whose equation is X; =0, we have Vpar= Ho OH, 0... 0H, But QAH= (Vp VO DOH= Var DU Q-10M, Thus IQAHI=1V,_- H+ but Va_¢ 0 Q'y-1=@, thus : IQAHl=1Vp_-OHI+1Qy_1 0H LOH = 1Vg-29Qy-1 0H; 1°) Suppose that H> Va_r- ‘Then, we have :1 Vq_p VHI=1 Vg: =1P*""(Fg)1= Ry. Furthermore, the linear form h defining H is such that h ¢ Fg[Xq-.X,_}- Indeed, if ha Baxi, ‘we have for alli21, Pj=(O:...:0:1:0:...:0) where the 1 is at the i" - coordinate, Pie Var and H> Var thus h(P) =0. But h(P) = a thus aj =O for all '2 1. Hence, 194-10 Hl=g QT AHL. ‘The quatric Q,_ 0H, of P*-7(Fy) is degenerate or no, acconding as H, is tangent or not 10. Q_1, Now: — IH, is not tangent to Q,_ , then by proposition 1, (since Q,_ 1m H, is nondegenerate in PT~2(F,)), we have : . 1Q 1 ABy =m 3# (Qe 1 1H) Da 2, Thus FQ Hla myth TF 11 QOH = y_2 + (XQ, OHg)— 1) g@-F- DP, — IFT, is tangent © Q,_ 1, then by theorem 1, we have 1Q 1 HAL = 3+ OQ. 1 VHA)— Dg" 9?, but by proposition 2 we know that @(Q,_ 1 0 He) = @(Q,__ 1).which is equal to «(Q) by definition, Finally, 1QaHl=m_.+ qf" (m3 +(Q-1) q*-7?) = Mq_2 + (Q= 1) 7-97. 2°) Suppose now that H' not contains Va Wehave Vag H=Hy Hy 0... 0H, 1 0H, thus 1Vq-2 OHI =IP®-F NR) 12 yp. 1h ¥ oiX; isthe linear form defining H, there exist necessarily one jr $j Sn, such that 4 #0. Thus QL (mo? ER 2Yet ow EY 182 Ya 1 oe 2a) © PMR) With Q,—1(%9--%e— 1) = Oand the x; are not all Zero }, ‘where tis such that BYE == 90K one Be 1 BE IB IG ‘Thus IQ aH g®@-*9-1Q 41 with Q,_ 1 a nondegenerate quadric of P'~ 1(F,), then :10 IQ, OHI =a" (2+ ((Q_1)- Dg?) and finally : (QA Hl = agp +98 (R02 + OQ 1)= qt?) Wy —2 + (OKQ)— 1) GFP, which concludes the proof.¢ 2.3, Intersection of two quadrics in P"(F,). ‘The subject matter of this paragraph is to estimate the number of points in the intersection of ‘two quadrics in PF) with n > 1. We give an exact value ofthis number in a particular case, and an upper bound in the general case (Theorem 3), inspired by an another upper bound of ‘WM. Schmidt (10] p.152). We need first a lemma : ‘Lemma : If Qy and Qo are two distinct quadrics in P*(F), then : 1 AQIsmq_1 +4"? Proof : By theorem 1, 1 Qy |= %y_1 + ((Qy) ~ 1) q@*~? if ris the rank of Qy. Thus = — if r2 4, we have 25=" 4. We set, fori equal I and 2 FX oXy) ‘The polynomials Py and Pp are not the zero polynomial (otherwise F and Fy would be 100), and are not also identically zero, since they have degree at most 2, and q > 4 implies that F and Fy have at most 2q°~1 4. ‘Thus it is possible to choose (1y..¢q) € Fy! such that Py (Cpr) #0. and Po(ChyCg) # 0. ‘Thus, after a nonsingular linear transformation and after divided by Py(cy Pa(Cnueg) Fespectvely, we may suppose without loss of generality that : Fy(KopenXq) = XG + Xo 81K en Kq) + B(KpyanXq) and F(X org) = Xp + Xo hy Keng) + Bg (KpoX)2 where gyhyeFgXpnXqly. and gyshze FelXty%al9 If we look at now the polynomials Fy and F, as polynomials in Xo, their resultant i a homogeneous polynomial R(X;,..Xq) of degree 4. By the well known properties of the resultant, we can say that for any common zero (in Fg"*!) (Kopx ro%g) Of Fy(XovanXp) and Fy(KoynoKghs We Bave ROS yn¥g) = 0. I we apply the Serre bound (see § 21) othe resultant R, we obtain that the number of zeros in Fy" of RO yusXq) is $4 q2~! — 342-2, Moreover, for such n-uple the number of possibilities for xis at most 2, andthe forms Fy and F, are of degree 2, thus the total number of common 22708 (xpeusX,) of F, and Fp in Fos seq! 6q'2, And by the following usual equality : ” Na = 1+ (4-D Np) ‘where Nq(F) represent the number of zeros in A'™*1(F,) = F.™! of F and Np(F) the number of zeros in P"(F,) of F, we deduce : 3. Projective Reed-Muller codes of order 1 associated to a quadric Let Qbe a quadtic in P?(Fy) of rank r, decomposing in disjoint union ofits vertex Vp and of| Q’r- 1 where Q,__ is the nondegenerate associated quadric of P'~ 1(F,). We will apply the results of § 2.2 to determine the parameters of the projective Reed-Muller codes of order 1 associated to Q. Since these parameters vary according to the type of the quadric Q, we have to distinguish three cases. ‘Theorem 4 (parabolic case) : Let Q be a parabolic quadric of PF) of rank r #1. Then the projective Reed-Muller code of order 1 associated to Qis a code with three weights: gt t= g28=F D2 ya gt My g@-F- D2 yy ag! with the following parameters length = nq. 1.dimension = n + 1, distance = q?~1 — g(@-*- 2 ‘Theorem 5 (hyperbolic case) : Let Q be an hyperbolic quadric of P°(F,) of rank r. Then the Drojective Reed-Muller code of order 1 associated to Q is a code with three weights : mae gt= t+ g@— 9, won gt, wee git + q@n- 92 _ q@n-#-2)2 withthe following parameters : Jength = %__ 1 +q@"~9?, dimension +1, distance = q?=1, cry ‘Theorem 6 (elliptic case) : Let Q be an elliptic quadric of P*(F,) of rank r > 2, Then the projective Reed-Maller code of order 1 associated to Qis a code with three weighs : wash! = g28=P, wom ght, wya qh! ~ g@-82 4 q@n-#- 2/0 ‘with the following parameters : length = q_1 ~ q@"~9?, dimension = n + 1, distance = q"~! ~ g(@"-9?, Let us remark that we recover the results of J. Wolfman as a particular case of these results (Gee [13)), indeed he had considered the case of nondegenerate quadtics : his results correspond to the case where the rank r = n+1. Note that, here, the case H7 Vq._, is excluded, and then ‘we find only two weights for the hyperbolic and elliptic quadrics, but still three weights for the parabolic one, We recover also the results of LM. Chakravarti (see [1]): it corresponds to the ‘case where the rank r= n, ‘Proof : The lengths of the respective codes are equal to the number of points of the respective Fq@ as introduced in § 1 defining ihe projective Reed-Muller code of order 2 associated to the quadric Q has for domain the vector space of quadratic forms ‘over Fq ; this is why we gave previously some results on the intersection of two quadrics of PPE). ‘Theorem 7 (parabolic case): Let Qbe a parabolic quadrc in P'(F,), m2. If q2 8 then the projective Reed-Muller code of order 2 associated 1 Q has the following parameters: aa eet length = fy dimension = 98+) dance 2h-!~ qt? Sy ‘Theorem 8 (elliptic case): Let Q be an elipic quadvic in PM) of rank r > 2. If q2 8 then the projective Reed-Maller code of order 2 associated to Q has the following parameters: length = x_1 ~ q@-7, dimension = 222) , ant oe distance 2 g?-"~ g@-?_ 6g8-?4 We reserve the case where the quadric is hyperbolic of rank 2 for the theorem 10 (we have indeed more precise results) ‘Theorem 9 (hyperbolic case of rank r 2 4) : Let Q be an hyperbolic quadric in PF.) of rank +24, Ifq2 8 then he projective Reed-Muller code of onder 2 associated wo Q has the flowing perevetas: length = nq_1 + 422-92, dimension = +2), on distance 2g?" 9-9 6gh-24 OT ‘Let us remark that we can have, for the theorem 9, the same results with a weaker hypothesis ‘on q when the rank of Q is equal to 4 or 6, namely q > 5. Now we consider the case of maximal quadrics, that is hyperbolic quadrics of rank 2. By the corollary of theorem 1, the numberof points of these quadrics reaches the maximum number of points ofa quadric, and it isin this sense that we call hem "maximal". We can remark that they are particular quadrics (they are the union of two distinct hyperplanes). The codes which are associated to them have a minimum distance precisely known. These codes will have a generalization in the next paragraph, ‘Theorem 10 (yperbotic case of rank = 2): Let Q be an hyperbolic quaric in PYF,) of rank 2. The projective Reed-Malle code of order 2 associated to Q has the following parameters Jength = n_ y+ 4°"! dimension =2@* 9) -2¢q-1) Proof : The length of the codes is the number of points of the quadric Q, and is given by ‘Theorem 1. Let Fe F(X Xalp and Q' = Zya(F), Q= Zpa(F). Either F and F are proportional, and then Q = Q. Remark that there is q — 1 such non zero forms F ; thus there is at least q quadratic forms vanishing in Q, hence in the kemel of the map ¢ defining these codes. We claim that there are no other forms in Ker(c), and thus the dimension of this codes is : FyXomXale (ng y(n mdi 0) = dima = PA BO*D _ hop (1Kex(0)1) 2+ (n+ 2 a2 ian _ nfo +3) = Genera jaan as Indeed, suppose now that F and F are not proportional, we have by Theorem 3: n-1 Gg gt-2 1QAQ sm. - Sa, 16 ~~ if Q is parabolic (Th 7), we have eee wy 2 + GH - SET 0e0028, Moreover, F and F cannot have a common factor of degree 1 since Q would be the union of two hyperplanes and thus would be hyperbolic. ‘The minimum disance follows from te ame inequality ofthe Theorem 3. ~ifQis lip (TH 8), Fand F cannot alo havea conmon factor of degree 1, and we have : n=l Gt my ~ SE 8 forr = 4, and thus a fortiori for r 2 4, ie. since r is even, r > 2. if Q is hyperbolic of rank 2 4 (Th 9), the same reasoning gives a fortiori the results (indeed the hypothesis q2 8 holds for more “smallest quadrcs). =i Qis hyperbolic of rank =2 (Th 10): + itherF and F have common facto of degree 1, and by the Theorem 1QNQ'1=m__+9"~? which is <1QI=x,_y+q"~!. + or F and F have not a common factor of degree 1, and by the lemma preceding Theorem 3 wehave 1QAQ’1s: {-? whichis <1 Qt. 1QU~ Cmq_ata?-?) Fy" defining the code is obviously one to one since d 1) is a code with thre weights : wi=(h = 1) qt), w= m3 and with the following parameters : length = mq_2+hq"~1, dimension =n 1, distance = (h~1) q°~?. hg?! +(1 — by gh? References om Chakravarti ILM., Families of codes with few distinct weights from singular and ‘non-singular hermitian varieries and quadrics in projective geometries and Hadamard difference sets and designs associated with two-weights codes, Coding. ‘Theory and Design Theory - Part I: Coding Theory IMA vol. 20. a Delsarte P., Goethals J.M. and Mac Williams F.., On generalized Reed-Muller ‘codes and their relatives, Inform. and Control 16 (1970) 403-442. a) Games R.A. , The Geometry of Quadrics and Correlations of sequences, IEEE ‘Transactions on Information Theory. Vol. IT-32, No. 3, May 1986, 423-426. io} 6 fa, m (8) a 10] ay 12) 03) v7 Hirschfeld J.W-P., Projective Geometries over Finite Fields, Clarendon Press, Oxford, 1979. Kasami T., Lin S. and Peterson W.W., New generalization of the Reed-Muller codes - Part I : Primitive codes, IEEE Trans. Information Theory IT-14 (1968), 189-199. Lachaud G., The parameters of projective Reed-Muller codes, Discrete Mathematics 81 (1990), 217-221. Manin Yu. and Viadut $.G., Linear codes and modular curves, Itogi Nauki i ‘Tekhniki 25 (1984) 209-257 J. Soviet Math. 30 (1985) 2611-2643, Primrose E.J.F., Quadrics in finite geomeiries, Proc. Camb. Phil. Soc., 47 (1951), 299-304. Ray-Chaudhuri D.K., Some results on quadrics in finite projective geometry based on Galois fields, Can. J. Math., vol. 14, (1962), 129-138. ‘Schmidt W.M., Equations over Finite Fields. An Elementary Approach, Lecture Notes in Maths 536 (1975). Serre, J-P., Lettre a M. Tsfasman, 24 juillet 1989, Yournées Arithmétiques de Luminy, Astérisque, S.MLE,, Paris, to appear. ‘Sorensen A.B., Projective Reed-Muller codes, to appear, Wolfman J., Codes projectifs @ deux ou trois poids associgs aux hyperquadriques une géométrie finie, Discrete Mathematics 13 (1975) 85-211, North-Holland.Decoding Algebraic-Geometric Codes by solving a key equation Dirk Ebrhard* 1 Introduction ‘The recent work about the problem of decoding Algebraic-Geometric Codes has led to an algorithm (e.g, see [2,3])- Another algorithm has been given by Porter, see (6,7,8,9,10), generalizing Berlekamp’s decoding algorithm. ‘The main step is to solve a so-called “key-equation”, For this purpose, Porter gave a generalization of Euclid’s algorithm for functions on curves, Unfortunately, therefore he had to impose some strong restrictions to the code and its underlying curve, such that the resulting algorithm works only for a very small class of Algebraic-Geometric Codes. Recently, the generalized Euclidian algorithm was investigated and corrected by Porter, Shen and Pellikasn ((11}) and Shen (12). Here, we will show how to generalize Porters ideas to all Algebraic- Geometric Codes and moreover, how to solve the key equation by simple linear algebra operations. Two observations on Porter’s methods have motivated our work: 1, The operations done by Porters algorithm at the so-called “resultant-matrix”, may be considered as a Gaussian algorithm, applied to the transposed matrix. 2. The key equation may be viewed as linear. ‘The result is given in section 2: A decoding algorithm of complexity order O(n"), that corrects up to [(d"—1~g)| errors, exactly as the well known algorithm does, In section 3 we describe how strongly both algorithms are connected. We conclude with section 4, giving a short overview over Porters algorithm and explaining, in what manner Porters work embeds in ours. 2 The decoding procedure 2.1 The code We will use the notations of {I}: Let X be a curve of genus g (i.e. a non-singular, abso- lutely irreducible projective curve defined over the finite field IF,), Py,...,Pq rational “The author is with the Mathenoatisches Institut IV der Heintich-Heine-Universtit, 4000 Dioeldorf 1, Germany. ‘The contents of tha paper are also part of the author's Ph. D. Thess ({13}) 19 points on X, and G a divisor which has support disjoint from the P,’'s, We will as- sume that 29 ~ 2 < degG@ Sn+g~1 and define D:= P, +...+ Py. Then the code C = C"(D,G) is the image of the linear, injective map Resp: 9(G-D) — Ft ” + (Reenny...,Reorn)- This is a linear fn, ky dl, ~ code with k > n—1—deg G-+g and d > d* = deg G— (29-2). For details see (1]. Note, that Resp may be extended to (X) in a canonical way. 2.2 A theorem for preparation Let @" be a divisor with support disjoint from the P,’s such that G—G" is effective and dim L(G’) = 0. An easy consequence is: 9(G — D) C 9G" — D), which we will use to generalize (7, Theorem V.1, p. 16]: ‘Theorem 1 There exists « vector space V: MG —D) CV C MG" D) such that Resply: V —+ IF} is an isomorphism Proof. Since Resplaia-p) is injective, it suffices to prove that Resplaar-p) is surjective. We have ker (Resplyo-p)) = {7 € MG" — D) : Resp, = 0, ‘Therefore -n} = (6) rank (Resplo’-p)) = dim 9G" — D) ~ imac’) = 9~1-deg(G! ~ D) ~(9 —1— deg’) deg D =n, using the Riemann-Roch-Theorem and the fact that 0 < dim L(G’—D) < dim Z(G’) = 0. ‘The problem of decoding is: For an arbitrary given y € IFS, find c € C with minimal Hamming distance wt(y—c). According to Theorem 1, Resp gives a correspondence of vector spaces: Q(G-D) c V c Q@'-D) Jrmo raw ocr If IF} + V, w ++ ne denotes the inverse map of Resply, then we can describe the decoding problem in terms of differentials: For a given ny, find ne € 9(G —D), such that ne 2= ny ~ ne has minimal number of Poles in Pyy...,Pq, ice. such thet m may te written as a fraction of functions with low degree. We will precise that in the next section.2.3 ‘The key equation Let y = e+e € IF? with c € C and F an arbitrary divisor on X. Let D, denote the unique Divisor with 0 < D, < D and P, € supp De # ev # 0, that is, De = ((ne)oo)hurp D+ Definition 1 By solution of the key equation we will denote any tripel (B.w.a) € (LE) \{0}) x (GF) x YG = D ~ F), satisfying Bn, = a+ w. Proposition 1 If deg F + wt(e) <4", then any solution (B,w,<2) of the key equation satisfies end ne = &. Proof. Let (B,w,a) be a solution of the key equation. From w = By, — a and 7, = e-t me one concludes w — Bn, = By.~a. If the two sides of this equation do not vanish, ‘we may estimate their divisor (w= Bae) 2 min((w),(B) + (m)) 2 min(G" FG'-F-D,)=G'-F-D, (Bye ~ @) > min((B) + (n.),(a)) > min(-F + G@- D,G-D-F)=G-D-F Since D and G’ have disjoint supports and G < G’, we get, (2 ~ Br.) 2 max(G' - F- D,,G-D-F)=G~D.-F, but deg(G—D, ~F) = deg @—wt(e)—degF > deg @—d" = 2g—2, what contradicts the ‘assumption that w—Bre is a non-vanishing differential. Therefore w—Bne = Bne—a = 0, what proves the statement, Proposition 2 If deg F > wt(e) +9, then there exists « solution of the key equation, Proof. By the Riemann-Roch-Theorem, dim L(F ~ D,) > 1 + deg F ~ deg D. - 9 > 1+wi(e) +9 —wt(e) — 9 = 1, which guarantees the existence of B € L(F ~ D.)\ {0} C ‘L(F)\ {0}. Now we have Bn, = Bn. + Bne, (Br.) = (B) + (1) > —F+@~ D, and (Bry — Bn.) = (Bne) = (B)+(n) 2 ~F+D.+G'—D, = G'— F, therefore (B, Bn., Bne) is a solution of the key equation, Corollary 1 Let F be a divisor of degree [2421]. Por any y € I such that there i €€ C with wile s= yc) < [EF2|, there exists o solution ofthe key equation. On the other hand, for eny solution (B,a.u) the equality 1, = % holds. Proof. One easily checks, that we have deg F+wt(e) < d* and deg F > wt(e)+9. Now apply Propositions 1 and 2 ‘The proofs of the Propositions show, that already the assumptions dim 9G — D, — F)=0 resp. dim L(F ~ D,) > 0 suffice to guarantee existence and uniqueness of solu- tions. This coincides exactly with the assumptions needed by the well-known algorithm. (see [3}). 21 2.4 Solving the key equation Assume to to be non-negative (i.e. d’ > g) and F to be a divisor of degree [£36] = tog. Ror an arbitrary y € FY consider the near map 6 : KF) + 4G'-D-F), Bow Bey Notice, that 9G — D ~ F) n 9(G" — F) = 0G ~ F) = {0}, since deg(G — F) = deg G — deg F > deg G ~ d" = 2g ~ 2. Hence there exists a vector space W such that Q(G'- D-F) =9(G - D- F/M" - Few. @ Let m, aor) denote the natural projections onto W resp. 2(G' — F). If there is any codeword c with wt(y—c) < to, then, by the Corollary, there will exist Be L(F)\ {0}, aE GDF) and w € 1G — F) such that 5,(B) = By = atu. Every such triple will suffice 1, = . In this case, the following algorithm will compute the error vector e: 1, Compute the matrix describing 5. 2, Determine B € ker(nw 0 6) \ {0}. 3. Compute w := raor-m(6,(B)). 4, Compute e := Resp §. If too many errors have occured, then either ker(nw6,) = {0} or the computed e won't suffice the conditions w(e) < to and y~e EC. 2.5 Realisation and complexity Let (¢s)rain denote the canonical basis of IF*. If the matrices describing &, are computed before the algorithm starts and once forever, the matrix of 8 = Shay Yobe may be computed with complexity order O(n®) in run time. Then parte 2 and 3 of our algorithm may be done by simple linear algebra operations; the first one with complexity order O(n*) and the second with order O(n2). We will now describe shortly how to realize the remaining third part in O(n?) steps: Assume w and B have been determined; now we want to compute Resp¥ for any Pc supp D. Let tbe a local coordinate on X around P, j m= —vp(P) and w = Soot! dt and B = 5” By" the local power series around P. Now,2 # Compute By, Bjsrs--. until By, # 0 occurs. # Compute aij-1 # Set Rese Since w aud B ase given as linear combinations of certain bacie differentials (sexp. functions), any coefficient of their power series may be computed as an corresponding linear combination if the power series at every point of supp D of any basis differential (resp. function) is known a priori. The complexity to compute one coefficient that way is O(n). Altogether, there have to be computed Soc1414 un(B) +n (F)) < 2n + deg ((B) + Flee < 20+ deg F corflicients, ence one can get alongwith complexity order O(n) in the decoding lgo- ‘ithm fourth part, Furthermore, an analogous computation shows that not more than deg F coeficient of each power sree around each peint ofthe bass cfferetils (erp. functions) must be knowa 8 prior 3 Essentially that’s nothing new ‘To demonstrate how our method of decoding is connected to that of [3] resp. [4], we will show that the main steps of each of the decoding procedures are equivalent. We first consider the map $:0(G! — D - F) + L(G — FY’ = {linear ¢: L(G ~ F) + F,}, where ®(n)(f) = Cia Rese, (f -1)- Proposition 3 If the supports of F and D are disjoint, then Shy is an isomorphism Proof. Since equation (1) holds, it suffices to show that 1, @ is surjective, 2. 2(G' — F)@ MG — D = F) Cc ker@ and 3. dimW =dim L(G - FY’. L: The map L(G — F) + If, f - (f(Pi),.--.f(Pa)) as kernel L(G - D ~ F) = {0}, hence is injective, therefore it suffices to prove that #:2(@’ — D- F)-> CE)’, S(n)(0) =F Resa.u, is surjective. Let (e¥),a1.» denote the canonical basis of (})Y. For any v, 9(G' — F — P,)\ 0(G" ~ F) is not empty 09 # consequence of the Riemann-Roch-Theorem, and the image by @ of any such differential is e Hence & is surjective, hence ®, too. ‘Propaition 3 Together with decomporition (1) show the exactness of 0+ (6 - D- F)@NG'~ F) + G'- D-F)* 4(G- FYY +0, that may be also derived in a canonical way of a short exact sequence of sheaves on the curve X. For details see [13] 2 2: If'n € 9(G' ~ F) then for any v, Resa = 0, hence #(n) = 0. Now take € (G = D— F) and f € L(G—F). Then fy € 9(~D) and, by the residue theorem is: DResn (fn) = 0. dimW dim(G' — D - F) ~ dim9(@" - F) - dima - D - F) = 9-1 deg(G! — D~ FP) (91 ~ deg(G’ — F)) - ~(9~1~deg(G-D =F) 1-9 + deg(@ - F) dim L(G - F) = dimL(G-F)’, by the Riemann-Roch-Theorem, using the fact that deg(@ — F) > 2g 2 and deg(G—D =F) <0. In our decoding procedure, the fundamental step isto find an element of ker(nw 05,)\ {0} which is, by Proposition 3, equivalent to finding something contained in Leap ° rw 08,)\ {0}. But (Bh o-mw 06(B))(f) = 8(5,(B)\f) = O(n, - BL) DRese,(f my B) DAP) B(P,)- Resa (ny) = LAP) BP) -w, for any B € L(F) and any f € L(G—F). Hence ly 0 mw 0 5, is exactly the error locating map E, in [4]. Finding a non-trivial element of its kernel is the main step in the known algorithm, too, 4 Porter’s decoding algorithm ‘We will give a very short overview over Porter's algorithm and explain why we consider ‘he presented algorithm as its generalisation. Porter makes some restrictions to the codes he treats, We will formulate them using the notations of Section 2: 1, G'= —P.,, where Pa is a rational point on X, distinct from supp D. 2. G is lineary equivalent to (deg G)- Par, G+ Pay is effective and deg G > 0. 3. (29 ~2)Pa. is canonical Especially assumption 3 won’t be realizable for most curves, but fortunately thie re- striction may be dismissed easly. First remember the isomorphism Resp: V —+ IF? of Section 2, and let €1,...4é, basis of V such that Resp(¢s),...,Resp(q) is the canonical basis of IFS. Furthermore,24 let 7 be a rational function? with (7) = G — (deg @) Pay, wo 0 differential with (wo (29 — 2)Pas, and m := —vp,(7). For any “received” vector y = e+e € 7, Porter defines a rational function called “syndrom” by Note that $ € R= UsoL(tPue), Swo = my (mod 7), ie. Sup = ny +f for some £ ER, and ~vp(S) 0 for some “small” divisor F). ‘This is the main idea of the method of decoding described in Section 2. References [1] J.H. van Lint, G. van der Geer, Introduction to Coding Theory and Algebraic Geometry. DMV Seminar, Band 12. Birkhiuser Verlag 1988. [2] J. Justesen, K.J. Larsen, A. Havemose, ILE. Jensen, T. Hgholt, Construction and decoding of a class of algebraic geometry codes. [EBE-IT 35(4)(1989), pp. 811-821. {3] A.N. Skorobogatov, $.G. Vladut, On the decoding of algebraic-geometric codes. IEEE-IT 36(5)(1990),pp. 1051-1060. {4] R. Pellikaan, On a decoding Algorithm for Codes on maximal curves. IEEE-IT '95(8)(1989), pp. 1228-1232. [5] S.G. Viddut, On the decoding of algebraic-geometric codes for q > 16. IEBE-IT 36(6)(1990), pp. 1961-1963, [6] S.C. Porter, Decoding Codes arising from Goppa’s construction on algebraic curves. ‘Thesis, Yale University, 1988. [7] S.C. Porter, Decoding Geometric Goppa Codes. Preprint. {8] S.C. Porter, Euclid’s algorithm, resultants and rational function representation on algebraic curves with a single point at infinity. Preprint. [9] 8.0. Porter, Dense representation of affine coordinate rings of curves with one point at infinity. Proceedings of ISSAC-89. [10] S.C. Porter, An efficient data structure for rational function on algebraic curves. Preprint, (11] S.C. Porter, B.Z. Shen, R. Pellikaan, Decoding geometric Goppa codes using an extra place. Preprint Bindhoven University, September 1992 [12] B.Z. Shen, Subresultant sequence on a Weierstrass algebra and its application to decoding algebraic-geometric codes. preprint Eindhoven University, May 1991 [13] D. Bhrhard, Uber das Dekodieren Algebraisch-Geometrischer Codes. Thesis, Diisseldorf University, 1991. “TAs a consequence, one doesn need assumption 2On the Different of Abelian Extensions of Global Fields G. Frey, M. Perret, H. Stichtenoth ©. Introduction Let q be a power of some prime number p, and let F, be the field with ¢ elements. Coding theorists are interested in explicitly described function fields over Fy having a large number of F,-rational places (or, equivalently, irreducible complete smooth algebraic curves over Fy with many F,-rational points). For small values of the genus, such function felds are often abelian extensions of the rational function field F(z). For instance, this is the case for Hermitian curves, some Fermat curves, and some Artin- Schreier extensions of F(z). Moreover, one way to exhibit families of function fields E/E, of genus growing to infinity and having good asymptotic behaviour (i.e., the ratio (umber of rational places/genus) has a limit > 0), is to construct a tower of function fields Ey CE, C Ea... over Fy, each step Eiz,/E; being Galois with an abelian Galois su. In other words, solvable extensions may have a good asymptotic behaviour, ef. (3). One aim of our paper is to show that abelian extensions E;/F (where F is some fixed function field over Fy, and Fy is assumed to be the full constant field of F and all iyi 2 1) are asymptotically bad (i.c., the ratio (number of rational places/genus) tends _ to 0 as the genus of i/F, goes to infinity). It should be pointed out that our method uses only elementary results from Hilbert’s ramification theory, ef. [2,4], and the finiteness of the residue class fields. In the case of Slobal fields, one may also use class ficld theory in order to obtain some results of this paper. 1, Hilbert’s Ramification Theory for Locally Abelian Extensions In this section, we consider the following situation. K is some field, 0 C K a discrete valuation ring and p Co the maximal ideal of o. Let L/K be a finite abelien field eztension with Galois group G (i.e. L/K is Galois, and its Galois group G is abelian). Let OC 1 be a discrete valuation ring of L with o C O and maximal ideal P, hence p= PNo. Throughout section 1, we suppose that O is the only discrete valuation ring of L containing o. Let k := o/p and I= O/P denote the residue class fields of o resp. ©. Then I/k isa finite icld cxtension, and we slall always assume that I/k ts separable, We choose some P-prime element x € P (i.e, P is the principal ideal generated by x), and consider the groups Ge {e€Gloz=z mod P forall re 0} and, for i> 1, {0 € Goo: x mod PH), ar It is well-known that the definition of Gi is independent of the choice of x, and GD Gy 2G D... 2 Gq = {1} for sufficiently large n > 1, see [2,4]. The factor groups Pi/P+ (for i> 0) are considered as vector spaces over I via (e+?)-(@+P)= zat PH? (20,0 P'), and G acts on P*/P*+! by 7a PHY) = ra) +P (in order to see that this action is well-defined observe that © is the only extension of o in L, hence r(P) =P for all r € G). We set a+ PH CPP |(a4PH!)=atP™ forall EG). Clearly, Xj is a k-subspace of P'/Pi?. Proposition 1: The dimension of Xj as a vector space over k is at most one. Proof: By Hilbert’s ramification theory 1/k is a normal field extension. Due to our assumption I/k being separable we obtain that 1/k is Galois. Moreover, any automor- phiam rin the Galois group of /k is induced by some 7 € Gi. m(=-+P) = r(2) +P for any 2 +P € O/P = | see [2]. In order to prove the proposition we can assume that ks # (0) Nochoste's + PH eX, wih e © PAPO. Since Pippi is a one-dimensional vector space over I (this is obvious) we have for all a; + PH € Xi a1 +P = (¢+P)- (a+ Pi) for some ¢ € O. For any 7 € G, the following holds: (6+ P)-(a+ PHY) = ay + PH = olay + PH) =1(e+P)-1(a+ PH) = 1(e+P)-(a+ P), 1s ¢-+ P is invariant under any auto- consequently r(¢-+P) = +P for any + €G. Thus c+ P is invariant morphism of I/k, ie. c+ P € k. This proves Proposition 1. . It ia well known that the mappings [G1 >t {Se ee (810 Be are embeddings of Go/G1 into the multiplicative group of I (resp. of Gi/Gi41 into the additive group P'/P*), and the definition of y and ¢ is independent of the choice of the prime element x. For our purposes, the following refinement is essential:28 Proposition 2: Under the hypotheses of this section, the image of ¥ is contained in *, and the image of yj is contained in X; for any i > 1. Proof: (a) Let o € G. We have to show that ro(¥(o)) = 9(c) for all 7» in the Galois group of 1/k. As before, 7 is induced by some t € G, and we obtain Wi(o) = (SE +7) = HED 4-p= ee) (we have used that G is abelian and r(x) is a P-prime element as well). (b) An analogous argument proves that yi(a) € X; for any o € Gi. . We recall some facts from ramification theory. Let f = f(P | p) = (I: k] denote the residue class degree and e := e(P | p) the ramification indez of P over p, ie. pO = P*. Since O is the only extension of o in L and Ik is separable, we have ef = [L : K]. Let s:= char(k) be the characteristic of the residue class field and g, 42 0. Then G1 is the unique s-Sylow subgroup of Go, and gy = ¢. The extension P | p is said to be tame if 9, = 1 (hence (s,¢) = 1), otherwise P | p is wildly ramified +P = (0) Let WC k* be the group of roots of unity in k. If W is finite, we set w= #W. Corollary 3: In addition to the hypotheses of this section, suppose that k contains only finitely many roots of unity. If P | p is tame then e < w. Proof: Consider the map ¥ : Gp —+ 1° as before. Since Gy = 1, is a monomorphism, By Proposition 2, the image of # ia contained in W. . In order to obtain a similar estimate for the ramification index also in the ease of wild ramification, we introduce the following notion: an integer i > 0 is called a jump (for Pl p)it Gs # Gita. Corollary 4: In addition to the hypotheses of this section, assume that k is a finite field. Then e < (##k)" where r denotes the number of jumps, Proof: © = 90 = (so/$1)(o4/42)-+-.+(gn/n41) where n is chosen such that gag = 1. Proposition 1 and 2 yield gi/gi+1 < #k for any i > 0. The corollary follows immediately.« Hilbert’s formula [2,4] states that the different exponent d 4=S@.-»). ‘This formula can be restated as follows. We consider the set {r4,-..,v-) of jumps (where 0S v4 dre 423 where r denotes the number of jumps. Proof: by the Hasse-Arf theorem, . 2. The Different of Abelian Extensions of Global Fields Tn this section, F denotes a global field. This means that either F is a number field, or F is an algebraic function field of one variable over a finite field F, (we assume that Fy is the full constant field of F). A place of F is the maximal ideal of a discrete valuation ring of F. If p is a place of F, its corresponding valuation ring will be denoted by op. The residue class field op/g is a finite field, and in the function field case we have Fy C op. The degree of p is defined by deg p = log #(op/p) (im the number field case, log is taken with respect to the basis e = 2,718. function field over Fy, we take log = log, - the logarithm with respect to the basis q). ‘The definition of the degree is extended to divisors of F (a divisor is a formal finite sum of places) by linearity.20 Let E/F be an abelian extension of F and Gal(E/F) be its Galois group. If p isa place of F, there are g = 9(p) places P,,...,P, of E lying over p (ie. pC Py). All of them have the same ramification index e(p) := e(P, | p) and the same residue class degree ) = f(Pr | p), and we have e(p)- f(p)- 9(p) = [E : F]. For an extension P = P,, of p in E/F, we consider the decomposition group G(p) = GP | p) = {o € GaE/F) | oP =P} (this is independent of the choice of the extension P since E/F is abelian), and the decomposition field Z = Z(p), ie. F C ZC E and G(p) = Gal(E/Z). ‘There exists the unique mazimal unramified subestension F CM C B. This means ea Places ‘of F are unramified in M/F, and M is a maximal subfield of E with this property. Let Ss= S(E/F) be the set of places of F which are ramified i it is well-known that $ is finite). , see BIR Lemma 6: With the notations as above, we have Lloe elo) 2 tog oes : Fl—logiM : F]. Proof: For any p € S, let Go(p) C G be the inertia group of i = $, let Gal) inertia group of p, see [2,4]. Its order is e(p), and if U C Gis the subgroup of G generated by all Go(g) (with p € nN ‘then M is the fixed field of U. Therefore ord U = [E: M] = [E: F\/IM : F. Since @ is abelian, ord US TT ord Gp) = J] elo). ves oes ‘Taking logarithms yields the assertion of the lemma. fe Let D(E/F) be the different of E/F. The main result of this section is the following: Theorem 7: Suppose that E/F is an abelian extension of global fields and FCM © E is the maximal unramifed mbestnsion. Inthe fantion Sld ase we ‘sume, in addition, that E and F have the same constant feld F,.. ‘Then the degree of the different D(E/F) satisfies “om ane deg D(E/F) > 31: F](loglB F] ~loglM : F). Proof: For p € S, let d(g) be the different exponent of a place P of E lying over p, and r(p) be the number of jumps, ef. section 1 (observe that we can ) be the nu , ef. section 1 (observe that we can apply the results of section 1 if F is replaced by the decomposition field Z(p)). We obtain at deg D{E/F) = SY Ato) des? PES Fie = Dal) dlp) J) dee Fd 2 Date)-r06)-e(6)-M06)-deep (by Propesition 5) ‘a le: F]- Sor()-degp te (by Corollary 4) > HB: F)-(loglE FI—logiM:F)) (by Lemme). Abelian Extensions of Function Fields Are Asymptotically Bad We want to prove a slightly more general result than we announced in the introduction. For an algebraic function field /F, (with Fy as its full constant field) we set enusof E sumber of rational places of E/Fy. 1 E/F is a Galois extension with Galois group G, we let G" be the commutator subgroup of G. The fixed field E** 3 F of G' is the mazimal abelian extension of F contained in B. In particular, if @ is abelian, G’ = {1} and E*} = B. ‘Theorem 8: Let F/F, be an algebraic function field and (E,)y>1 be a sequence of extension fields of F with the following properties: (Fy is the full constant field of F and all Ey. (i) E,/F is Galois with Galois group Gy. (ii) ord (G,/G4) — 00 a8 v—+ 00. Then the quotient N(E,)/9(Ey) tends to zero as v —+ 00. Proof: There is a constant h (the class number of F) such that any abelian unramified extension M/F with the same constant field Fy is of degree [M : F] < h, ef. [1]. We consider the maximal abelian extension F, C By of F contained in E,. By (ii), the degree ny i= [Fy : F] —+ 00 as v —+ 00 , and the degree dy, of the different D(F, /F) satisfies the estimate82 4, > Ln s(logny —logh) by Theorem 7. The Hurwitz genus formula yields (Fe) 2 nsCalF)-1) + Ze 2 ns(a(F) —14 Hlogny —logh). On the other hand, we have the trivial estimate N(F,) —D, together with the zero function, and 92(D) will denote the Fy- ‘vector space of all rational differentials on X, defined over Fy, with divisor (n) > D, together with the zero differential. Put 1(D) = dimp,1(D) and i(D) = dimp,9(D). ‘The Riemann-Roch Theorem states that W(D) = deg D+1-9+i(D). Also, we will write Do (resp. Dac) for the divisor of zeros (resp. divisor of poles) of D. Hence we have Do > 0, Dey > 0, (Supp Do) N (Supp Dao) = 0, and D = Do - Deo 'V. D. Goppa (3,4) realized that one could use divisor theory on a curve to define nice codes. A g-ary linear code of length n and dimension k is a vector subspace of dimension k of F?. The minimum distance of a code is the minimum number of places in which two distinct codewords differ. For a linear code, the minimum distance is also the minimum weight of a nonzero codeword, where the weight of a codeword is the number of nonzero places in that codeword. A linear code of length n, dimension k and minimum distance d is called an [n,k,d}-code. Let C denote an [n,k,d-code over Fy. A generator matrix of C is a k xn matrix whose rows form a basis for the code. A patity check matrix for C is an (n —k) x n matrix B of rank n—k such that AB' = 0 for some generator matrix A of C. ‘Two eades C and C* are called dual if a generator matrix of C* is a parity check matrix of C. The code C has minimum distance d if and only if in any parity check matrix for C every d~ 1 columns are linearly independent and some d columns are linearly dependent. “The first author was supported by the Alexander von Humboldt-Stiftung while visiting Universitat GHS Basen within the GMD-CNPgq exchange program. ‘The sec- ond author was partially supported by a grant from the Louisiana Education Quality Support Fund through the Board of Regents.34 Let G be a divisor on X defined over Fy and let D = P, + P;-+----+ Py be another divisor on X where Pi,...,Pq are distinct Fy-rational points and none of the P; is in the support of G. The geometric Goppa code C(D, G) (cf. (6) is the image of the linear map a: L(G) + F} defined by SF (F(Pi), S(Pa)s-y f(Pa))- ‘The geometric Goppa code C*(D,G) (cf. {6]) is the image of the linear map a* : 9(G — D) + F} defined by 1+ (resp, (n),te87,(7),-.-,te8p, (7))- ‘The codes C(D,G) and C*(D,G) are dual codes ((6, 11.3.3]). Alternatively, if one fixes ‘a nonzero rational differential « with (canonical) divisor K;, then, the image of a” is the same as the image of the map * : L(K + D~G) = F, defined by Fr (resp, (feo), resp, (fu),... resp, (fu). We will be mainly interested in the code C*(D,G). This code has length n and dimension k= dim L(K + D -G) —dim (K -6). In particular, if G has degree greater than 2g~2, then k = dim L(K-+D~-G). If k > 0, then the minimum distance d of C*(D,G) satisfies the well-known bound 42 deg(G) — 29 +2. 2. Now let P denote a (closed) Fy-rational point on X and let B be a divisor defined over Fy. We call a natural number 7 a B-gap at P if there is no rational function f on X such that (A) + Ble =P. We say that 7—1 is an order at P for the divisor K — B (where K denotes a canonical divisor on X) if we have K-B~(7-P+8, where E> 0, P ¢ Supp(E), and ~ denotes linear equivalence. It follows from the Riemann-Roch Theorem that 7 is a B-gap at P if and only if 7 ~ 1 is an order at P for the divisor K — B. Note that with this terminology, the usual Weierstrass gaps at P are the O-gaps at P. Tet 7 and 74 be B-gaps at P. Put G= (45+ -)P +28. @ Our divisor D used to define the code C*(D,G) will be of the form DaPit Patt Pay where the P; are n distinct Fy-rational points, each not belonging to the support of G. 35 (2.1) THEOREM. Assume that the dimension of C*(D,G) is positive (where G is Sosea specified in (1)). Then the minimum distance of C*(D,G) is at least deg(G) - 24+ 3. PROOF. Put w i exists 1 € 2(G — D) with exactly w simple poles Pi, Pas. have leg G— (29~2). If there exists a codeword of weight w, then there Pw in Supp(D). We then (n)o 2 Go and (1)oo S Goo + Pr + Pe +--+ + Po =29- lows = Gy and Hence, 2g—2= deg(n) > deg Go — deg Gao - w = 292. It follows that (n)o = Go (her Gat ht Prt - + Py. Thus there exists a canonical divisor K of the form — (Pt Pptet Pe) ® K= Since 74 is a B-gap at P, we have K-Bx~(%-WP+E, (3) where B > 0 and P ¢ Supp(B). From (1), (2), and (3), we have B+ (Pit Pate t+ Pe)—(B+75P) ~0. “WP, contradicting the ‘Thus there exists a rational function f such that ((f) + B). fact that 7; is a B-gap at P. a (2.2) REMARK. If B= 0, then deg = 1y +74 ~ 1 and it is natural to ask which integers can be written as the sum of two Weierstrass gaps. In this direction, G. Oliveira [7] showed that if X is nonhyperelliptic, then each r, for 2 4(P)}, where I(P: p(X) + H) denotes intersection number. ‘Assume from now on that the linear system [K — B] is base-point-free. If 1(P) < ‘n(P) <--+ < qwaa(P) are the B-gaps at P, then for the morphism associated to the linear system | — Bl, we have 6(P) = ai(P) - 1.36 (2.3) THEOREM. With notation as in Theorem (2.1), assume moreover that 7341 = “q+ and that the osculating space LY-")X (with respect to the morphism defined by 1K — B)) does not meet the curve ¢(X) at another F,-rational point. If C*(D,G) has Positive dimension, then its minimum distance is at least deg G ~ 29 +4. PROOF. By Theorem (2.1), the minimum distance of C*(D,G) is at least deg G—29+3. Put w = deg G — 2g + 3 and suppose that there is a codeword of weight w. Proceeding ‘as in the proof of Theorem 1, we see that there is a canonical divisor K of the form K=G+Q-(A+Pit-- +P), @ where Q is an F,-rational point on X. Using equations (1), (3), and (4), we get B~—1yP-Q+ (Pit Pte + Po) +B. Since yj and 7y++1 are both B-gaps at P, we see that Q ¢ {P, Pi, Ps,...,Pu}U Supp(E). Hence, UB +45P) # (B+ 43P +Q). By Riemann-Roch, we then have LK - B-%4P)=1(K - B-P-Q). This equality means that for any divisor A linearly equivalent to K ~B, if A > 14P, then A> 7;P+Q. Now, ¢;(P) = Ye-1 = % soil A 2 ej(P)P, then A 2 5(P)P+Q. It follows from this that Q is in L3~"X, contradicting our hypothesis. » (2.4) REMARK. Suppose B = 0; ic., g: X —+ P?l is the canonical morphism. Ifat an F,-rational point P, we have 7, = 29 ~ 1, then L{?-*)X, the osculating hyperplane at P, meets X only at P and hence LX, for j = 1,2,...,g — 1, meets X only at P, 80 the condition on the oaculating space in Theorem (2.3) will be satisfied. ‘The next result gives another useful criterion for verifying this osculating space condition. (2.5) PROPOSITION. Suppose that B = 0 and that X is not hyperelliptic. For J€{1,2,...,9—2}, we have 41 = 25 => (LYM X) NX = {P}. PROOF. First we remark that 7341 < 2j for j = 1,2,...,9—2. For suppose 741 > 2). ‘Then there are at most j gape less than or equal to 2j. Hence the dimension of L(2jP) would be at least 2j+1—j +1, contradicting Clifford’s Theorem. Now, the intersection divisor of X and LY~”)X is of the form A = (7j41—1)P +E, where E > 0 and P ¢ Supp(E). By the geometric version of the Riemann-Roch ‘Theorem (1, p-12], we have dim L(A) = deg A ~ dim LE- x = deg A (j - 1). ® From Clifford’s Theorem, we have dim L(A) < (deg A)/2+ 1. @) or From (5) and (6), we get deg A < 2 —1, which implies that deg F is at most 2) — 9341. So, if 7441 = 2, then B =0 and the Proposition follows. 4 We note that Proposition (2.5) was proven by G. Oliveira [7] in the case j (2.6) EXAMPLE. We give an example to show that the converse of Proposition (2.5) is false, and that onc can sometimes deduce the fact that LY-” .X = {P} by testing the condition 141 = 21 for some ! > j. Let X denote the nonsingular curve with function field Foa(z,y) where y*+y = 23. If P = (1,q), with a +a = 1, is an Feq-rational point, then by Garcia-Viana [2] the gap sequence at P is 1, 2, 3, 4, 5, 10, 11. Taking ! = 5, we see that “ius = 10 = 2l; hence, LY) X meets X only at P. It follows that L‘)X must also meet X only at P, but notice that for j = 4, we have 741 =5 4 2 =8. In the case where the next t consecutive integers after 7; are also B-gaps, one can prove results similar to Theorem (2.3) by assuming more conditions on the osculating spaces. (2.1) THEOREM. Suppose that 744 = 1 +t for t = 1,2. Suppose that LY)X does not meet X in another F,-rational point. Also suppose that L2-YX does not meet any of the following lines: (i) A line joining two other Fy-rational points. Gi) A tine joining two Fye-rational points that are conjugate over Fy. (iil) A tangent line at another F,-rational point. If C*(D,G) has positive dimension, then the minimum distance of C*(D,@) is at least deg G29 +5. PROOF. Put w = degG—29+4. Since LX does not meet X at another F,-rational point, the same is true for L-Y.X. Thus by Theorem (2.3), the code C*(D,G) has rinimum distance at least w. Suppose that there is a codeword of weight w. ‘Then there is a canonical divisor K of the form K=G+A-(A+ Pits +Pa), @ F A= Qi + Qe of degree 2 defined over Fy. From (1), (3), and for some positive di (7), we have Br —ayP—(Q1+Q2) + (Pit Pate: + Pu) +B, (8) where E> 0 and P ¢ Supp(E). Now, either Qs and Qs are both Fy-rational points or they are F,e-rational points that are conjugate over Fy. Since 7; +2 is a B-gap at P, we cannot have Q1 = Q2 = P. By applying the proof of Theorem (2.3) to the divisor G+ P (note that 741 and 44341 are B-gaps), we see that neither Q; nor Qs can equal P, since then the other point would lie in LX. Also, we have that neither Q; nor Qq is in {Pi,Pa,...,Po}U Supp(E).. It then follows from (8) that we have UB +P + Qi) # UB +P +Q+%), as38 where these are now vector spaces of rational functions defined over F,». It follows from. the Riemann-Roch Theorem that UK - B-4;P —Q1) = (K - B~4;P -Q1-Q:). ‘Thus for any divisor H ~ (K — B), we have H2GP+Qi=6(P)P + Qi =e H> P+ Qt Qr=6(P)P+Q+Qr. ‘Therefore, the osculating space LY-"'® meets the line joining Q, and Qo, contradicting our hypothesis. 4 ‘The conditions on LS in Theorem (2.7) can be expressed as: LY-X misses all the lines determined by degree two F,-rational divisors A, where P ¢ Supp(A). By using induction, one can generalize Thoorem (2.7) in the case of ¢ consecutive gaps after +4; to obtain the following result, (2.8) THEOREM. Suppose that 741 = 71) +1 for some natural number t. Suppose that for each s € {0,1,...,t—1} the osculating space L3-!**)X misses the linear space spanned by each degree ¢ — s Fy-rational divisor with support disjoint from {P}. If C*(D, G) has positive dimension, then its minimum distance is at least deg @— (29 — +041). 3. We present some examples illustrating the theorems in the previous section. (8.1) EXAMPLE. Goppa [4, pp. 139-145] considers the plane quintic yz + yz! 4. 2° ~ 2323 over the field F. This curve has 17 points over Fy. Put P = (0,1,0). ‘Then the gap sequence at P is shown to be 1,2,3,6,7,11. Goppa shows (p. 144) that the code C*(D,16P), where D is the sum of the remaining 16 F,-rational points, has minimum distance at least 8. ‘This follows from Theorem (2.3). Our proofs of Theorems (2.1) and (2.3) are essentially generalizations of Goppa’s argument. (3.2) EXAMPLE. Let X denote the curve y + y = xt defined over Fy, where 2 5. This is an example of a Hermitian curve. There are q* + 1 F,z-rational points on X, the maximum possible (by the Hasse-Weil bound) on a smooth curve of genus 9 = (q~1)4/2 over Fa. Let P be an Fy-rational point on X and let D denote the sum of the remaining q° F,s-rational points. The codes C(D,sP) have been studied extensively by Tiersma [11] and Stichtenoth [9], and recently Yang and Kumar [12] have determined the exact minimum distances of these codes. In this example, we show how to use Theorem (2.7) to give an alternate derivation of some of the results of Yang and Kumar. Let P = (a,6) be an Fye-rational point on X. The canonical morphism y : X —+ ‘P#* given by a Hermite basis at P of the regular differentials (cf. (2,8) is o((z,y)) = PR4s (2a) BOs (ea): ), (y~b)~ a¥(x—a) is the tangent line at P. We note that P meets X only at P (with multiplicity ¢+1), Denoting by Pzy the point on X at infinity, we have that 3: (e-a)Pr; 39 +1). By (8} (or see (2), the Weierstrass nongaps at P smaller 0 a+) 24,24 +1,2(9 +1) (a-4)a,(¢~4)q-+1,.-.5(9-4)(g+1) (9~3)a,(9—-3)q+1,...,(¢~-3)(9 +1) (q-2)a, (4-2) +1,...5(9-2)(+1) (a-4(a+1) +1. Note that 7,741, and 7 +2 are Let Gx denote the divisor (1)-+ 72 -1)P = (14 +(q—4)(q+1))P and let D denote the sum of the other g? F,z-rational points on X. ‘The usual lower bound on Goppa codes would show that the minimum distance of C*(D, Gy) is at least deg Gi — (292) = ‘th ~ 2(q+ 1). We claim that the conditions of Theorem (2.7) are satisfied here, so that the minimum distance of C*(D, Ge) is at least 4 ~2(q +1) +3. ‘To establish this claim, first note that LY. 9X = {P}, since 7 = 29 - 1. Next, we will see that L2-X = L2~X (ie,, the osculating space of codimension 5 at P) does not meet any line determined by two (distinct) Fy-rational points (different from P). Since g is given by the Hermite basis associated to the point P, the osculating 9X in Pol is given by Xp-1 = Xp-2 (y-8 = 0 (where the X; are homogeneous coordinates). Suppose (21,11) and y(22, 2) are distinct points on ‘(X), both different from (P). The line determined by these points is {ap(e1,n) + By(e2,1n) :0,0 € 5}. Put a P= Peisu) = (8) ~a%(2;—0) fori =1,2. If there is a point on the above line that also lies in L{¢-®X, then we have Pe + ph (») a(21 — a)PE + B(x, — a) PE oP + phy? Note that P, # 0 and Fy # 0; otherwise, there would be another point on the curve (the point (2, 34)) lying on the tangent line to the plane curve X at the point (a,0). Multiplying the first equation in (+) by Py and comparing the resulting equation with the third equation in (+), we have Py Multiplying the frst equation in (+) by (21 ~a) and comparing the resulting equation with the second equation in (s), we see that 25 = 23, From Py = Py and 2; = 2a, we may conclude that yy, = yy40 ‘This contradicts the assumption that p(z1, 91) and (22,92) were distinct points, One argues similarly if one of the two points (2y,y:) is Pa. Finally, suppose (21,1) is an F,s-rational point distinct from the point (a,2). In order to determine the tangent line to g(X) at (1,91), we take derivatives with respect to 2: Zeeun)=af aad 2 fer,n) = (e109 Put F; = P(z1,y1). From the definition of g, the tangent line to p(X) at g(z1,y) is {aylzrn1)+6V: 0,6 € Fy), where ¥ = © (a-3)(2-a)*PF* : PE + (q—3)(e1 a) AF: (q—2)(01 a) PF), ‘Suppose there is a point Q on this tangent line that also lies in LY’-X. Since no rational point on X except (a,2) lies in this osculating space, we see that 6 0 at Q, 80 we may assume # = 1 at Q. Then at Q we must have + (q— 3)(21 — a)ePe* = a(x; — a)PY + BE + (q— 3)(21 — a) B+ = 0 ‘Multiplying the first equation by (z1 — a) and comparing the resulting equation with the second equation, we see that F, = 0. But this says that (1,9) belongs to the tangent line to X at (a,0), a contradiction. Again, one can argue similarly if (z1,y1) is Pao. Since the projective change of coordinates (z : y : 2) ++ (x: z: y) induces an ‘automorphism of X that takes (0: 1 : 0) to (0: 0: 1), we see that the conditions of Theorem (2.7) are also satisfied at Pay. Put dy equal to the minimum distance of C*(D, Gy). If we take 7, = (2+m)q— 1, where 1 < m < q~3, then by Theorem (2.7), we have dy > mg. Put z = (2~a1)(t—a2)-+-(2—ay), where G;, 02,...4dyy are distinct elements in Fs. Then the differential d2/z has a zero of order (q?—q—2)4'mg > deg Gy at Pa, and has exactly mq simple poles, so itis a codeword of weight mq. ‘Thus, the minimum distance of C*(D,Gs) will be exactly mg when % = (2+ m)q—1. Using Riemann-Roch to compute the dimension of this code, we see that C*(D, Gy) is a [ea +4 —4(q+2m ~ 1)/2,mg}-code when = (2+ m)q 4. Although our theorems show that the minimum distance of the code C*(D, eP) is related to the Weierstrass gap sequence at P, we close with an example to demonstrate that this minimum distance does not depend solely on the gaps. Specifically, we give a curve with two points with the same Weierstrass gaps such that corresponding Goppa codes have different minimum distances, (4.1) EXAMPLE. Let X denote the nonsingular plane quartic at pyzt nts? 4229 4 20y2z = 0 a“ over the field Fs. ‘Then X has the following eight Fs-rational points: (0,1,0), Pa = (3,11), Py = (41,0) P, = (3,21), Po = (4,8,1), Po = (2,41) P=(0,0,1), and P! = (2,3,1), ‘The points P and P' are (ordinary) flexes of X, so the Weierstrass gap sequence at each of these two points is 1,2,4. Put D = P,+Py4---+Pe-+P! and D! = Py+Prt---+Pe+P. ‘The functions 1 and z/x are a basis for L(3P) and a generator matrix for the code (0, 3P) is biiaridide 024243 3)" It can then be seen that the weight enumerator (cf. [6, p. 38]) of C(D,3P) is 1+ 1225 + 42° +827, By the MacWilliams identity (cf. [5, p. 39]), the weight enumerator of the dual code C*(D,3P) is 1+ 122? + 8025 + 40024 + 80425 + 11802° + 64827. ‘The functions 1 and (x + z)/(y + 22) form a basis of L(3P’). A generator matrix for C(D',3P") is then biidiiada 0301033 ‘The weight enumerator of C(D',3P*) is then 148244425 41227, ‘Thus C(D,3P) and C(D',3P") have different minimum distances. However, the weight ‘enumerator of C*(D',3P") is 1+ 242? + 602" + 36024 + 9242° + 10802" + 6762", and so C*(D’,3P") has the same minimum distance as C*(D,3P). But now consider the divisors 5P and 5P”. The functions 1, 2/2, and (2+ y)2/2? form a basis for L(5P). A generator matrix for C(D, 5P) is The weight enumerator of C(D,5P) is then 1+ 82° + 3625 + 642° +162" and the weight enumerator of the dual code C*(D, 5P) is 1+ 3225 + 5224 + 15625 + 27226 41122",42 ‘The functions 1,(z + 2)/(y +22), and (x + 2)(2 + 32)/(y +22)? form a basis for USP’). A generator matrix for C(D',5P’) is yaaiidida 0301033). o104202 ‘The weight enumerator of C(D’,5") is then L429 + 162 + 3225 4 402 4 3227 and the weight enumerator of C*(D',5P’) is Ltda? +1229 + 7224 417625 + 25228 + 1082", In pasticular, the minimum distance of C*(D,5P) is 3, while the minimum distance of C*(D', 5P’) is only 2. REFERENCES [i] B. Arbarello, M. Comalba, P.A. Griffiths, J. Harris, Geometry of algebraic curves, Volume I, Springer-Verlag, New York, 1985, P| A. Garcia and P. Viana, Weierstrass points on certain non-classical curves, Arch. Math. 46 (1986), 315-322. 8] V. D. Goppa, Algebraico-geometric codes, Math. USSR-Izv. 24, no. 1, (1983), 75~ 91 [4] V. D. Goppa, Geometry and codes, Kluwer, Dordrecht, 1988, {5] J. H. van Lint, Introduction to coding theory, Springer-Verlag, New York, 1982. (6) J. H. van Lint and G. van der Geer, Introduction to coding theory and algebraic geometry, Birkhiuser, Basel,1988. IM] G. Oliveira, Weierstrass semigroups and the canonical ideal of non-trigonal curves, ‘Manus. Math. 71 (1991), 431-450. {s] F-K. Schmidt, Zur arithmetischen Theorie der algebraischen Funktionen Il. Allge- meine Theorie der Weierstrasspunkte, Math. Z. 45 (1939), 75-96, [9] H. Stichtenoth, A note on Hermitian codes over GF(¢#), IBEE ‘Trans. Inform. ‘Theory 34, no. 5 (1988), 1345-1348. {No} K.-O. Stabr and J.P. Voloch, Weierstrass points and curves over finite fields, Proc. London Math. Soc. (3), 52 (1986), 1-19. [11] HJ. Tiersma, Remarks on codes from Hermitian curves, IEEE Trans. Inform. The- ory, IT-33 (1987), 605-609. £N2] K. Yang and P.V. Kumar, On the true minimum distance of Hermitian codes, these Proceedings. On a Characterization of Some Minihypers in PG(t,q) (q=3 or 4) and its Applications to Error-Correcting Codes Noboru Hamada? ‘Tor Helleseth? Department of Applied Mathematics Department of Informatics ‘Osaka Women’s University University of Bergen Sakai, Osaka, Japan 590 N-5020 Bergen, Norway Abstract A set P of f points in faite projective geometry PG(t,¢) is an {f,m;q)—minihyper if m (= 0) is the largest integer such that all hyperplanes in PG(t,q) contain at least m points in F where ¢ > 2, f > 1 and q is a prime power. Hamada and Deza [9} [11] characterized all {202 + 29943, 2%e + 299it,q}—minibypers for any integers t,q,@ and B such that q 2 5 end 0 0. Recently, Hamada (5), (6] and Hamada, Helleseth and Ytrehus (18] characterized all {201 + 2en,200 + 2vist,9}—miaihypers forthe case ¢ > 2 and q € {3,4}. The purpose of this paper is to characterize all {2va+41 + 2ug41,2ve + 2ug;,q}—minihypers for any integers t,q,0 and f such that q € {3,4},0 2 and q is a prime power. If |F H| > m for any hyperplane H in PG(t,) and IP | = m for some hyperplane H in PG(t,q), then F is called an {f,m;t,q}—minkhyper whore m 2 0 and [Al denotes the numberof elements inthe Set A. In the special case t = 2 and m > 2, an {f,m;2,q}—minihyper F is also called an m-blocking set if F contains n0 1—flat in PG(2,q). ‘The concept of a minihyper (or a minehyper) has been introduced by Hamada and Tameri (19). actly supped by Gratin fo Sete Resear he Minny of Bau, Seno end Cute wncr Cotas Nebr x. a20HN2 2 Pil appre by he Scns Jp Sutava Fonte, Prially supped byte Norwegan Rasa Coun for ‘Sexes he Hoon“a In the case k > 3 and 1 < d < q+ — q, d can be expressed uniquely as follows: ‘ eot- (ade) using some ordered set (¢,41,442,-++,4) in U(k — 1,9) where U(t,q) denotes the set of all ordered sets (¢,11,11,+++,44) of integers e,h and yy auch that () 0 < © 0. & ps (@) Characterize all {en + Z masnete + suit} minihypers in the case where there exist such minibypers. ~ “ Let A(tq) be the set of all ordered sets (A1,d2,---sq) of integers and Xi (F=1,2,-+-50) such that 1 2 and A; = Aj for some distinct integers i and j. In the case 9 = 2 and dy = Az, Problem 1.2 was solved by Hamada [7], In the case 1= 8, Problem 1.2 was solved completely by Hamada (6], [7], Hamada and Deza [8] and 46 ‘Hamada and Helleseth (12}-(17]. In the case = 4, q > 5 and dy = Az < Ay = As, Problem 1.2 was solved by Hamada and Deza (9), [11]. Recently, Hamada, Helleseth and Ytrehus [18] solved Problem 1.2 for the case = 4, 4 € {8,4} and (21, 424s, 44) = (0,0,1,1) (Cf. Propositions 11.3 and IIT.4 in Appendix). ‘The purpose of this paper is to solve Problem 1.2 for the case» = 4, q € {3,4}, At = Az < As = Ae and As # Ai-+1 using the results in Appendices J, IK and Ill, The ‘main result is as follows. ‘Theorem 1.1, Let t,q,a and 6 be any integers such that ¢ = 3 or 4,0 26 +1, F is a {2va41 + 20g41,2ve + 2vg;t, q}— minihyper if and only if F is a union of two a—flas and two A—fiats in PG(t,q) which are mutually disjoint, Remark 1.1, Hamada and Deza {9}, [11] showed that in the case q > 5 Theorem 1.1 holds for any integers ¢,a and such that 0 2042, Corollary LL. Let n = vy — 2va41 — 2upyr and d = gt) — 29 — 298 where g = 3 o40 5, F € 7(0,0,2,3;t,¢) for any {201 + 20,200+ 2ozit,q}— minibyper FP. ‘Theorem 2.2, Let 1, and q be integers such that t > 28 > 6 and g = 3 or4. If F is a {2v2 + 2up41,201+ 2vg;t,q}— minihyper such that (a) |F' G| = 2ug_ for some (#2) flat @ in PG(t,¢) and (&) F's € F(0,0,8 — 1, ~ 1;¢,9) for any hyperplane Hy (1S 5 Sa+1) in PG(t,q) which contain G, then F € F(1, 1,8,.;t,9). Remark 2.1. Since F(1,1,/,8;4,9) = 0 in the case = 26 (ef. Remark I.1), Theorem 2.2 shows that in the case ¢ = 28, there is no {2u2 + 2vp41,2m + 2ug;t,q}— minihyper F which satisfies the conditions (a) and (b) in Theorem 2.2. (Proof of Theorem 1.1) It follows from Proposition J.1 and Remark J.1 that if F ¢ F(a,<1,8,B5t,4) in the case t > 28 +1, then F is a {2ve41 + 20g41, 200 + 2vp;t,9}— minthyper. Conversely, suppose there exists a {2va+1 + 20941, 2ve + 2ug;t,q}— minihyper F for some integers t,q,a and 6 such that g = 3 or 4,0 (8-1) +1, FH; € F(8—1,8 —1;4,9) for vyged and FN Hy € F(0,8—1,8~1;4,9) for j = a,q-+ 1. Hence it follows from () and Proposition 1.3 (¢ = 2, #1 = 2 = ) that (1) in the case ¢ < 28 — 1, there is n0 {201 + 2v941,200 + 2v9:4,4}— minibyper F which satis the condition (a) and (2) inthe caset > 20, F € F(0,0,8,8;t,4). Since ¥(0,0,8,8;t,9) # 0 if and only ift > 28+1 (cE Remark I.1), there is no {2x + 2vg41,2v0 + 2ug_13,q}— minibyper in the case t = 28. (B) In the case 6 = & = fy = 0 and 8,41 = 2, it follows from Remark 1.2, Proposition 1.5, Case I and induction on f that (i) in the case t — 1 < 2(8 — 1), there is no {2vp,2up_1;t,9}— minihyper in Hj, a contradiction, and (ji) in the case 4-1 > 2(8—1)+1, POH: € F(B ~ 1,8 ~ 1;t,q) fori =1,2,-++,g and FA Hys1 € F(0,0,8 - 1,8 -154,9). Hence it follows from (), Proposition 13 and Remark I.1 that (1) in the ease t < 28, there a7 is no {2v; + 2up41,209 + 2ug;t,9}— minihyper F which satisfies the condition (b) and (2) in the case t > 28-+1, F € F(0,0,8,85t,0). From (A) and (B), it follows thet Theorem 1.1 holds in Case Il - and x3) It follows from Proposition 1-2 (¢ ) that FA Hy is a {201 + 2ug,200 + 2og. 5 +)q-+ 1. Hence it follows trom Remark 1.2, Cases 1 and il that (i) in the case t= 1 < 2(8 ~ 1), there is no {201 + 20g,209 + 2vp_1;t q}— minibyper in Hh, a contradiction, and (i) in the case t — 1 > 2(8 — 1) +1, FAH; € F(0,0,8 ~ 1,6 — 1;t,4) for j = 1,2,+--,g+ 1. Using Theorem 2.2 and Remark J.1, it ean be shown that Theorem 1.1 holds in Case I. Case IV : (ax2 and Bra+2) It follows from Proposition 1-2 that F 0 H; is a {2vq + 20p,2vq-1 + 2up-1;¢,q}— minihyper in Hj for j = 1,2,-++,g-+1. Hence it follows from Remark 1.2, Cases I, Il, I, induction on a and f, Proposition 1.3 and Remark I. that Theorem 1.1 holds in Case IV. This completes the proof of Theorem 1.1. 3. The proof of Theorem 2.1 Let F(A, 92,-++,Anits gy 705151543), FA(0,0,1,15¢,3) and F(0, 0,1,154,4) (¢ = 1,2,3,4) denote the families given in Appendix I, Definition 17.1, 11.2, II1.2 respectively. In order to prove Theorem 2.1, we prepare the following four lemmas whose proofs will be given in Sections 5, 6, 7 and 8 respectively. Lemma 3.1. In the case ¢ > 4, there is no {201 -+2vs,.2v0 + 2v2;4,3}— minihyper F such that (a) [F'9 G| = 2 for some (t~2)—flat G in PG(t,3) and () FO Hh € F (1, 154,3), FN Ha € F(1,1;1,3), FO Ha € F(0,1,1;4,3) and FA Hy € F(0, 1, 15¢,3) where wo = 0, v1 = 1, v2 = 4, vy = 13 and the Hy denote four hyperplanes in PG(t,3) which contain G. Lemma 3.2, In the case ¢ > 4, there is no {20 + 2vs,2v0 + 2va;¢,3}— minibyper F such that (@) [F 9 G| = 2 for some (t—2)—flat G in PG(t, 3) and (®) FAH; € F(1, 54,3), PN Ha € FL 1;t,3), FO Hs € F(0,1, 15,3) and PA Hy € F(O,1, 13,3) where the Hi, denote four hyperplanes in PG(t,3) which contain @, Lemma 3.3, In the case t > 4, there is no {2m + 2vs,2v9 + 2v2;t,3}— minihyper F such that (a) |F’ 0 G| = 2 for some (t~2)—fiat G in PG(t,3) and (0) FAH € F(1,1;t,3), FO He € F(1,1;t,3), FA Hs € F(1,1;t,3) and FO Hy € Fo(0,0,1,1;t,3) for some 6 in {1,2,3,4} where the H; denote four hyperplanes in PG(2,3) which contain G. Lemma 3.4. In the case t 2 4, there is no {2m + 2vs,2vy +202;t,4}— minibyper F such that (@) |F 9 G| = 2 for some (t-2)—flat @ in PG(t,4) and (0) FAH; € F(1,1;¢,4) for i = 1,2,3,4 and F Hs € Fe(0,0,1,1;1,4) for some @ in {1,2,3,4} where v = 0, v1 = 1, v2 = 5, vy = 21 and the Hy denote five hyperplanes in PG(t,4) which contain G. Proof of Theorem 2.1) Suppose there exists a {201 + 2vs,2v0 + 2v2;t,q}— minihyper F for some integer t > 3. Then it follows from Proposition I.2 (¢ = 2, h = 2, 1 = #2 = 2)8 that @ [FG] = 2 for some (¢ — 2)-Bat G in POL) and () FAH; is a {65+ 20,2936,9}-mintyper in H, fo ny bypepane H's 3 < 4+ 1) a Pale) cf which conan G where te fare nomegnive integers sch that 6; = 2. Without of generality, we can assume that either (a); = by = «+» = by. =O and fy = bys = 1. 0 B) 6 = O and yes = 2. Case: (q23, 5, and 65 = 64 = 1) It follows from Rematic 1.2, Propositions 1.5 and 11.2 that () in the ease t—1 = 2, there is no {2en,20134,3}—minibyper in Hy, a contradiction, and (i in the ease t= 12 3, FH € F(1,134,3), PO Ha € F(1,354,3) and either FN Hy € F(0,1,2;8,8) of FN Hy € F(0,1,1;4,3) for i = 3,4. I follows ‘rom Lemmas 3.1 and 3.2 that F'M H; € F(0,1,1;4,8) for i = 3,4. Hence it follows from @, Proposition 1:3 (¢ = 2, h= 2, wi = 2 =2) and Remark 1.1 that Theorem 2.1 holds in Case 1, Case Is (qu4, f = b2 = by = 0 and & = b= 1). It follows from Remark 1.2, Propositions 1.5 and I1T.1 (a= 0,8 = 7 =1) that (@ in the case 1-1 = 2, there is no {202,2e1;t,4}—minityper in Ha contradiction, and (ji) in the case t-1> 3, FH € F134), for d= 1,2,3 and FH; € F(0,1,1)4,4) for j = 4,5. Hence i follows from (), Proposition 1.3 and Remark I.1 that Theorem 2.1 holds in Case I Case Il: (q=3, 6; ‘and 6 = 2) I follows from Rematk 1.2, Propositions 15 and 113 that () in the case ¢— 1 = 2, there is no {2vp,2013¢,3}—minibyper in Zs, 4 contradiction, and (ji) in the case 11 > 3, PO Hi € F(1,1;4,3), for i = 1,2,3 and cither F1 Ha € F(0,0,1,1;4,3) or F 0 Ha € (00,1, 13,8) for some 8 in {1,2,3,4). Hence it follows from (), Lemma 8.3, Proposition 1.3 and Remark I.1 that Theorem 21 holds in Case I Case IV : (g=d, 6: = 62 = by = 64 = 0 and bs = 2), It follows from Rematk 12, Propositions 1.5 and ITZ4 tat (inthe case -—1 = 2, there is n0 {2,204;4,4}—minihyper in H,, a contradiction, and (i) inthe ease ¢—1> 3, FH € F(L, 1it,4) for i = 1,284 and either FH € F(0,0,1, 13,4) ot PAHs € Fa(0,0,1,1;4,4) for some din {1,2,3,4). Hence it follows from (), Lemma 34, Proposition 13 and Remark [1 that Theorem 2.1 holds in Case TV, This completes the proof of Theorem 2.1. 4. The proof of Theorem 2.2 Lemma 4.1. Let t,6 and q be integers such that {> 28+1>7 andq>3. Fis 8 {202 + 20p41,201 + 2¥9;1,q}—minibyper such that P = X UY; U Ya for some set X of 2e2 points in PG(t,q)) and some Afats Yi, Ys in PG(t,2) which are mutually disjiat, then X € F(1,1;t4) and F € F(1,1,8,8;t,9). Proof, If |X H| > 2 for any hyperplane H in PG(t,q) and |X 9 H| = 2 for some hyperplane hyperplane H in PG(t,q), it follows from |X| = 2v2 and Proposition 1.5 that 2 X € F(1,15t,9) and F € F(1,1,8,6;t,9). Since IFN 8] =|XN8}+ DNB] and =] I%in B| = vp of upys for any hyperplane H in PG(t,q), it is sufficient to show that there 49 is no hyperplane H in PG(t,q) such that |F 0 H| orl + dup Suppose there exists a hyperplane H in PG(t,q) such that [FH] = 3+ 2ug, 4+ 29p, 29 + Ypvis 1+ ¥9-+ Ypvi Depys OF 14 Dopys Case 1: (PA M| = 801+ Bop) Suppose there exists (6~ 2)—fat G in HF such that [PAG| < 2op-a. Let Hi (i= 1,2,---yq) beg hyperplanes in Pt, 9), excep for H, which contain G, Since [P| = 2va + 2upq and [FP Hi] > 2 + 2ep for i = 1,2, 1% it follows: from qypca = up—1 and qyp = ogg —1 that [F| = [PHL + SUF 0 ~ LPG} 2 2ups1+29-+3 > [Fl a contradiction. Hence [FM G| > vy + 2vg-1 for any (t — 2)—flat @ inl. Using Proposiion 1.6 (0 =~ 1,01 = 1yep.1~ 2), we have [P| > on + 2up > 342yp = [FAM a contradiction. Therefore, there is 0 hyperplane HT in PG(tg) och that [FN] = 3 + 2ug. Case I: (ged and |P 1 H| = 4, + 209). Using a method similar to Case 1, it can be shown that there is no hyperplane HT in PG(t,q) such that [F' H| = 4-4 2op in the case g > 4, Case It: (gn3 and|F 0 H| = v2 209). Suppose there exists a (1—2)— fat G in H such that PG] 2ypqrt02+6 > |F|, a contradiction, Hence |FNG| > v1 + 2ug_1 for any (t — 2)~flat Gin A. Since |PB| = %9 + 2vp, it follows from Proposition .6 tat there exists a (¢ ~2)-fat @ in HT such that |F 1G] = vy + 2op-1. Let 52,3) be three hyperplanes in PG(t,3), ‘except for H, which contain G. = ug + upg, 1+ 0p + mma, Qopsr 2 Since Y [FA (04\G)| = |F| - [FN A] = 3(2-3°* +1) +1 and [FN (H,\G)| = WP tk] —]F 0 G| > 2-89-41 for = 1,2,5, there exists a hyperplane I in {I, Is} such that | = (20 + 2ug)-+1 = 34-2yp, a contradiction. Hence there is no hyperplane 2H in PG(t,8) such that [F 0 H| = 4+ 2vp in the case q = 3. Case IV : (F1H| = vp + ypia). Using « method similar to Case I, it can be shown that there exists a (t—2)—flat G in HT such that LP G| = vp_1-tup. Let I; (i = 1,2,-+-,4) be q hyperplanes in PG(t, 9), except for H, which contain G. Since 3} |PA(HA\O)| = LPI IPO B| = o(* +2) +2 and La (\@)| = PROM LNG] > Pt 49 for 6 = 142,595 there exists « hyperplane Tia {Ul1,Ha,-++, Tg} such that [FA T1| = 3 + 20g or 4+ 2up, a contradiction, Hence there {sno hyperplane H in PG(t,q) such thet |F 0H] = vp + vpaa- ‘Similarly, it ean be shown that there is no hyperplane Hin PG(t,4) such thet |F' 0 H1| = ups. Case VF H| = 14 vp + up4s oF 1 + 2ypy). Using a method similar to Case I, it can be shown that there exists « hyperplane II in PG(¢,q) such that [FN H| = 3 +2up, 8 contradiction. This completes the proof.50 (Proof of Theorem 2.2) Let F be any {202+ 2up41,201 + 2ug;t,q}— minibyper which satisfies the conditions (a) and (b) in Theorem 2.2 where t > 28 > 6. Then FA Hy = {Pa, Pa} U Wa U War for some points Px, Px and some (8 ~ 1)—flats Wis, Waa in Hi which are mutually disjoint for i = 1,2,---,q +1. Since |G W| = vg-1 oF vp for any (8 —1)—fist W and any (¢ ~ 2)—fat G in H,, it follows from LFAGI = Quer that Py 8G and GNWy = Vj (= 1,2--,04 1, j= 1,2) for some (9 ~ 2)-fat Yin. 7 + es Let X = U {Pa,Pa}, i = U Wa and Ye 1 Let Bj = Hi A (Wu @ Wa) for i = 1q+1. Then Ej is a (8 —1)—fiat in Hi such that GE; = Vi for i = 3,4,-++,q+1. Note that (2) Win U War UEaU+-U Egat is a A—fiat in PG(t, 9) and (i) either Win = Bi or Was Ei = Vi for each i, Hence if Wit = for i = 3,4,-++,0-+ 1, then Yj is a 6—flat in PG(t,9). Suppose Wir E; = Vi for some i in {3,4,++-,9-+ 1}. Without loss of generality, we can assume that Wa, = Ea, Way = Ba, ++) Wena, = Boi and Wo 0 Ep = West M Bysi = Vi for some @ in {3,4,---,9+1} where Bx = War. Let be a (¢~ 2)—flat in Hyan such that By41 C0, 1M Wosaa = Vi and 20 Va isa (8 — 3)—flat in G. Let Oy (1 = 1,2,-+-,q) be g hyperplanes in PG(t, q), except for Hy41, which contain 9. Then Il; Weis = 90 Weys,s = Vi and Ih, Way i a (8 — 2)—fiat in oy 19 Waal = vg-i) for P= 1,2,-++,9 and §= 1,2,+,9+ 1. Since Wi C Ha, War C a and War C Hp, for some integers a, 65, 64,-++ 545 in {1,2,-++s4}, there exists « hyperplane I in {U,,Io,-++, Ty} such that I Wa = Vi for i= 1,2,-++,9, Hence it follows from F = X UYiUYs that [FO = |X 00] +[%i 0M + Wat} = LO] + Mil + (q+ Lega = IX AM + (G+2)ypa < 2v1 + 2up unless 9 = 3,8 = 3 and [XN] = 8. Case I: (ge4 or q=3 and Pad). It follows that |F 0 Il| < 2v + 2vp, a contradiction. Hence Wa = Ey for i = 3,4,---,q-+1. This implies that Yi is a P—Auat in PG(t,q). Similarly, it can be shown that Yo is a 6—fat in PG(t,q). Hence it follows from Lemma 4.1 that P € F(1,1,8,5t,9) in the case ¢ > 28 + 1. In the case t = 28, there do not exist two A—flais Yi and Ys in PG(t,g) such that Yin Yo = 0. Hence in the case t = 29, there is no {2v2 + 2vp41,201 + 2up;t,q}—mininyper F which satisfies the conditions (a) and (b) in Theorem 2.2. Tis implies that Theorem 2.2 holds in Case 1. Case If : (q=3 and 3), It follows that 0 = 3 or 4, i.e, either (a) Wai 9 Es = War 0. Ea = Vi or (8) War = Ea and Wai Be = Vin (A) In the case War = Bs, we can assume without loss of generality that Wir C Th, War C Ui and Wa C Mh. Hence there exists a hyperplane Il in {Il2, Is} such that Pu ¢ WE and 1 Wa = Vi for i = 1,2,3. This implies that [X NII] < 7, ies [FU] < 2» + 2vp, a contradiction. Using a method similar to Case I, it can be shown that Theorem 2.2 holds in the case (A). (B) In the case Ws Es = Vi and Py, € Ey there exists a hyperplane It in {11, a, 13} such that Pay ¢ I and IM Wa = Vi for i = 1,2,3. This implies that |X NTI] < 7, [FAD] < 2x + 2vp, a contradiction, Hence Theorem 2.2 holds in the case (B). 51 (O) In the case Wai Ea = Vi and Par ¢ Es, let be a (t — 2)—fiat in Hy such that Pu ¢0, Bsc, AAW = Vi and QV, is a (f — 3)—fat in G. Let Uy (1 = 1,2,3) be three hyperplanes in PG(t, 3), except for Hs, which contain. Since Psy ¢ Th for! = 1,2,3, there exists a hyperplane II in {Mhy, la, Ig} such that [F 111] < 201 + 2vp, a contradiction, Hence Theorem 2.2 holds in the case (C). This completes the proof of Theorem 2.2, 5. The proof of Lemma 3.1 Suppose there exists a {2v; + 2vs,2vp + 2va;t,3}-minihyper F which satisfies the conditions (@) and (b) in Lemma 3.1. Then FAG = (Fi, Ps}, FN Hy = Lu U Lia FO Ha = In U Lan, FO Ha = La U Lan U {Ps} and FN He = V\{Q1,Q2,Q3,Q4} for some 2—flat V in Hy and some 4—arc {Q1,Q2,@s, Qc} in V where Diy and Lip are 1—fiats in Hy such that La M Lin = 0, GM Lin = {Pi} and GM Lin = {Pa} for i = 1,2,3 and Py is 4 point in H3\G and GV is a 1—flat in G which contains two points P and Ps. Without loss of generality, we can assume that GV = {Pi, Ps, Q1,Q2}- In order to prove Lemma 3.1, itis sufficient to show that there exists « hyperplane TI in PG(t,8) such that [FP Il| < 2v2 = 8 Let By = Hi (Im © Lai) for i= 1,4. Then is @ 1—flat in Hy; such that GB; = {74} for i = 1,4. Note that either Lu = Ey or Inn By = {Pi}. Case I: (Lu 0 By = {Pi}). Let M = GA (Lu@ Ei). Then M is a 1—fat in G passing through Pj. Let © be a (t ~ 3)—flat in G such that 591M = {Pi} and P) ¢ ©. Let (5 = 1,2,8) be three hyperplanes in PG(t, 3), except for Hy, which contain the (2) flat BOE; in Hh. Since Py € V, Pa € Lia, Po ¢ Ul; and M ¢ Ui; fori = 1,2,3 and j = 1,2,3, we have |V AHLj| = 4, [bia Hy| = 1 and La MM; = {Fi} for i = 1,2,3 and j = 1,2,3, Since Za C Ma, Loi C Ha and Py € Up for some integers a and 8 in {12,3}, there exists a hyperplane I in {Uh,1a,Ms} such that Py ¢ Wand IA Lin = {Fy} for i = 2,8. ‘This implies that there exists a hyperplane I in PG(t,3) such that [F Tl] = Zand + C0 We) A] $34 V0 = 7 < 209, contradiction. Hence Lyi = B,. Case H : (Li = Ey). Let B be a (t~ 3)—fiat in G such that Py € 5 and Py ¢ 3. Let 11; (j = 1,2,3) be three hyperplanes in PG(2,3), except for Hs, which contain the (~ 2)—fiat © @ Es in Ha. Since L1y C Ha, Io C Hay Za1 CHa and Ps € Up for some Integers and 8 in {1,2,3}, there exists hyperplane Tin {Mf Ma, Mg} such that Fy ¢ TE and 119 Li = {Pi} for i = 1,2,3, ie, [FM] = Land + (FH) | < 202, ® contradiction. Hence there is no {2v1 + 2vs,2v0 + 3on;t,3}—minihyper F which satsties the conditions (@) and (b) in Lemma 8.1. 6. The proof of Lemma 3.2 Suppose there exists a {2v1 + 2vs,209 +20n;¢,3}—minihyper F which satisfies the conditions (@) and (b) in Lemma 3.2. Then FG = {Fi,Pi}, FOB; = LU Lis, PO Ha = ImU Lag and FH; = VA\{Qj1, Qin, Qjs, Qj4} for some 2—flat Vj in Hy and82 some 4—are {Qj1, Qj2, Qjs, Qja} in Vj (J = 8,4) where Dix and Li are 1—flats in Hy such that Li A Lin = 0, G0 bat = {1} and G0 Lip = {P2} for i = 1,2. Since GN V; is a 1—fiat in G which contains two points F, and Ps, we can assume without loss of generality that G1 Vs = G11 Va = {Fi,P2,Q1,Qo} where Qj = Qi and Qin = Qo for j = 3,4. Since {Q:,Q2,Qis, Qua} and P, © Py are a 4—are and a 1—fat, respectively, in the 2—fiat V4, it follows that (Pi ® Fs) 1 (Gas ® Qua) = {44} oF {4}. Without loss of generality, we can assume that (P, @ Fa) (Qas © Qua) = {Fr}. Let E be a (t ~ 3)—fiat in G such that Pi € ¥ and P ¢ B. Let My (I = 1,2,3) be three hyperplanes in PG(t,3), except for H, which contain the (¢ — 2)—fiat B @ (Q4s © Qua) in Hy, Then [Lie Ohl = 1, [Lo2 0M] = 1, [¥s A Mil = 4 and Ve Th = {Qas, Qa, Pr, RY (ie, (F 1.V4) 9 Th = 2) for 1 = 1,2,3 where R denotes the point in Q4s © Qua except for three points Pi, Qas and Qu. Since Lr C Ha and Lm C Tg for some integers a and f in {1,2,3}, there exists ‘@ hyperplane I in {W1,1o,I3} such that 19 Lu = {Pi} and WA Ln = {Fi} ie, [Fan] = [200] + Lendl + (FAVA) ND] + (FAV) AD) -1 <7 < Boa contradiction. Hence there is no {2v1 + 2v,200 + 2va;t,3}—minihyper F which satisfies the conditions (a) and () in Lemma 3.2. 7. The proof of Lemma 33 Suppose there exists a {201 +2vs,20y + 2vy;t,3}—minibyper F which satisfies the conditions (¢) and (b) ia Lemma 3.3. Thea FAG = {F,P:}, FA Hy = InU Lu, POH) = Ln U Lan, FO Ha = Lu U Laz where Dix and Lip are 1~flats in Hi; such that Lan Lia =0, GLa = {Pi} and GN Lin = {Ph} for i = 1,2, Case I: (9=1). It follows from Definition 11.2 that PA Hs = V\{Q1,Q2,Qa} for some 2-flat V in Hy and some 3-are {Q1,Q2, Qa} in V. Let By = HaM(Li1 @ Lai) and let ‘bea (t~3)—flat in G such that P; € D and Py ¢ E. Let 11;(j = 1,2,3) be three hyperplanes in PG(G,3), except for H, which contain the (t ~ 2)—flat 3 @ Ey in Ha, Then there exists ‘ hyperplane II in {Ts, 2,13} such that |F T1| < 2v2, 2 contradiction. Case I: (0=2). It follows that F'n Hy = (V\S) U{P} for some 2—fiat V in Ha, some 4-are S$ = {Q1,Q2,Qs, Qs} in V and some point P in H4\V. Let Ey = Ha (Li ® Lai) and GOV = {Pi, Pe, Q1,Q2}- (A) In the case P ¢ Eg let M = GN(Es @ P) and let 3 be a (t~3)—flat in @ such that EMM = {Pi} and Pr gE. Let Il; (j = 1,2,3) be three hyperplanes in PG(t,3), except for He, which contain B @ Es. Since P ¢ M1; and |V A Mij| = 4 for j = 1,2,3, there exists a hyperplane TI in {U, Ta, 1s} such that |F'7Il| < 209, a contradiction. @) In the case P € Es, let N = GN (Ea @ Qs) and let E be a (t — 3)—flat in G such that N CE and Pp ¢ B. Let Hi (j = 1,2,3) be three hyperplanes in PG(t, 3), except for Hy, which contain £ @ Ey. Since VI; = P; @ Qs (.e., |(V\S) | < 3) for j there exists a hyperplane IT in {M1, Tp, lg} such that [F710] < 2ve, a contradict Case I : (03), It follows that F'0 Hy = LU K* for some 1—flat L in Hy and some ‘minihyper K* in ¥(0,0,1;¢,3) such that LOK* = 0, GAL = {Pi} and GNK* = {P}. Let 53 V be the 2~flat in Ha which contains K* and let M = GV. Let Ey = Ha A\(Li2 © Ian) and let 3 be a(t ~ 3)—flat in @ such that Py ¢ B and D7 M = {Ps}. Let Hj (j = 1,2,3) be three hyperplanes in PG(t,3), except for Hs, which contain © @ Es. Since M ¢ B and |K*NL*| < 3 for any 1—fiat L* in V, VU is a 1—fiat in V and [K*N(V O15) < 8 for j = 1,2,3, Hence it follows from [in Uj] = 1 and |G AUgl =1 (4 = 1,2,3, j = 1,2,8) that there exists a hyperplane II in {M1,, 112, 3} such that [PO] < 2e2, a contradiction, Case IV ; (624), _It follows from Definition 11.2 that Fn Hy = K for some minihyper K in Fa(0,0,1,1;4,3) where K = {(€o); (€:)s(2)s(€s)s(2€0+ 1), (20 + &2)s (260 + 3), (261 + &2), (2€1 + €3),(2€a + €5)} for some four linearly independent points (£o),(€1)s(€2) and (és) in PG(t,3). Let W be the 3—flat in Ha generated by four points (Eo), (E1),(2) and (6s) in PG(t,3). Since G isa (t—2)~fiat in the (~1)—flat Hy such that KG = (FN H,)NG = PNG = {P1,Pa}, WG must be a 2fat (denoted by A) in G such that KNA = {P,, Px}. Without loss of generality, we can assume that Py = (2g + €:) and Pi = (22 + 3). Let J = P@ Py Ai = J @ (261 +6), Ae = J @ (b), As = J @ (G2) and Aa = J (60+ 1 + £2 + &3). Then A; (i= 1,2,3,4) are four 2—fiats in W which contain J and KM Ax = {(2f0 + €1), (20 + €2), (20 + &s)s(2Er + €2), (261 + 3), (22 + &a)}, KMD = {(Eo)s(E1)s (2&0 + €:)s (26a + &a)}, KM As = { (a), (Es), (260 + £1) (2Eo + &)} and KM M4 = {(2f0 + £1), (22 + &3)} where A = Ay. Let By = Hg (Lu @ Lm). Then ‘Ba is a 1—fiat in Hg such that GOEy = {Pj}. Note that either Eq C W or Ean = {Fi}. (A) In the case Ey C W (ey Ea C A) for some é in {1,2,3}), there exists a 2—fiat V in W such that Ey CW, Pa ¢ V and |KAV| <4. Let N = GNV. Then N is a1—flat in @ such that Py € Nand Py ¢ N. Let be a (t ~3)—flat in G such that N CE and Pp ¢ E. Let I; (j = 1,2,3) be three hyperplanes in PG(t,3), except for Hs, which contain 3 @ Es, Since |Zie Uj] = 1 and WOU, = V fori =1,2,3 and j = 1,2,3, there exists a hyperplane II in {I,, 112, 13} such that [FAT] = 3+ |KAV| <7 < vp, a contradiction, (B) In the case Ex W = {Pi}, let V be a 2—flat in W such that P, € V, Pa ¢ V and |KAV| <4. Let M be a (t ~2)—fat in Hy such that By CM and QNW = V. Let 1, (5 = 1,2,3) be three hyperplanes in PG(t,3), except for Hs, which contain 9. Since WAM; = WOM = V for j = 1,2,8, there exists « hyperplane Il in {Ily, 12,103} such that LPO < 2va, a contradiction. This completes the proof. 8. The proof of Lemma 3.4 ‘Suppose there exists a {201 +20s,2vy + 2025t,4}—minihyper F which satisfies the conditions (2) and (6) in Lemma 3.4 where v = 0, m = 1, v2 = 5 and vy = 21. Then PNG = {Pi Pa} and FH U Mia for some 1—flats Max and Mez in H; such that Ma 0 Mia = 0, G0 Ma = {Pi} and GM Miz = {P2} for i = 1,2,3,4, Case I : (8=1). It follows from Definition 111.2 that F Hs = K for some minihyper K in F(0,0,1,1; 1,4). Let V be the 2—fat which contains K. Let B= Hs (Mi © Mai).54 ‘Then E is a 1—fit in Hg such that G0 E = {Pi}. Note that () KNG = PAG = (Pj, Fy} and (i) either ECV or ENV = {Pi}. (A) In the case EC V, let © be a (¢— 2)-flat in Hy such that AAV = EB. Let Tl, (I= 1,2,3,4) be four hyperplanes in PG(t,4), except for Hs, which contain 0. Since Pr € Mo, Pr€ V, Po ¢ 1 and Ph ¢ Th for i = 1,2,3,4 and 1 = 1,2,3,4, it follows that that [Mia Th] = 1 and VN = VN = B. Since Mir C Wey Mar C Hay Mai C Ts and ‘Mai C Ile for some integers «,6 and ¢ in {1,2,3,4}, there exists at least one hyperplane Win (MiyMa,Ms,Ta} such that 1 Ma = (Pi) for ¢ = 1,2,8,4, This implies tat \ranl= SiMe ntl + Fn He) <4+1V 9] =9 < 20, ‘contradiction, Hence Vn = {P. Note that 202 = 10 in Lemma 3.4. B) In the case V1 E = {Fi}, let be a (t— 2)-flat in Hy such that BC 9 and QNV = N where N is a 1—fiat in V such that Py € N and Pe ¢ N. Using a method similar to (A), we have a contradiction. Hence 6 # 1. Case It : (0=2 or 3). Using a method similar to Case I, it can be shown that there exists ‘ hyperplane II in PG(¢,4) such that [F' II] < 2vz, a contradiction, Hence 0 # 2,3. Case I : (G=4). It follows from Definition II1.2 that FA Hs = LU K* for some 1-flat L in Hy and some minihyper K* in ¥(0,0,1;¢,4) such that L1 K* = 0 where K* = {(wo); (wi); (w2), (wo + 01), (wo + w2), (wr + w2)(wo + or + w2)} for some noncollinear points (19), (wi) and (w2) in Hs. Note that |K*N E| < 3 for any 1—fiat E in V where V denotes the 2—flat generated by the three points (wn), (1) and (w2). Since |LNG| = 1 or 8 and (LUK*) NG = (FAH) NG = FNG = {PF}, we can assume without loss of generality that K*G = {Pi} and LNG = {Fe}. Le E = HQ (Mu @ Mz). Then either £ CV or ENV = {Fi} (A) In the case EC V, let 9 be a(t —2)-flat in Hy such that Py ¢ 0 and NV = E, Let My (! = 1,2,3,4) be four hyperplanes in PG(t,4), except for Hs, which contain 2. Then [Ma AI] = 1, |LATh| = 1 and VAI = E fori = 1,2,3,4 and I= 1,2,3,4. Hence there exists @ hyperplane Min {Ua,TayM,TI4} soch that \Fag| = 2 Mend + Gan + 1kene} < Vine = {hh (B) In the case VN E = {Pi}, let 9 be a(t —2)~fat in Hs such that EC Mand QV = N where N is a1—iat in V such that P, N and Pe ¢ N. Using a method similar to (A), we have a contradiction. Hence 0 4. This completes the proof of Lemma 3.4, (PN) NG = 8 < 2v2, a contradiction, Hence Appendix I, Preliminary results in the general case Let U(t,q) denote the set of all ordered sets (e,1,42,-++,pn) of integers e, h and pi: (§=1,2,-+-h) such that 0 < ¢ < q—1,1 <4 < (t-1)(g—1), 1 Sn Sma $+ Son 3 and 55 15d 0. Let Fu(e,#1,42,---,#hit,@) denote the family of all unions of ¢ points, a wi-Hat, a pe—flat, «++, a py—flat in PG(t,q) which are mutually disjoint where (¢,41,42,+++,4n) € U(t,g). ‘As occasion demands, we shall denote Fal pit anata) by Fs Ay ta) wie 9 = bbe, = OG 12e~ and Acts = 1 (5 = 1,2,-7+yh). For example, F(a,,7,5;t,9) denotes the family of all unions U Vi of an a eee flat V2, a y—flat V3 and a 6—flat Vy in PG(t,q) which are mutually disjoint where O< a y+ 1, then F is @ {ev1 + vp41,€v0 + v,5¢,q}—minihyper. UF € Folens aye«sspastig) inthe case h > 2 and t > ays + ya +4, then F isa {on + Semenov + 3 waite} minty. & a Remark Li. It is known (¢f. Theorems 2.2 and 2.3 in Hamada and Tamari (20) that Fu(essiy4ayssssnitsa) # 0 if and only if either (8) h = 1 or (b) h > 2 and #2 wnat mn +1 where (€,44154425-++s464) € U(t9). Proposition 12. (Hamada [3). If there exists a {enn + Zevnemt Sonitad —minihyper F for some ordered set (¢,1,12,---s 4) in U(t,q), then | FA [> ‘ 3 eat for any (¢~2)=at A in PG(t9) and | PAG |= BP for some (¢—2)—Bat Cha Ulta). Let Hy (JY 2,---49-42) bog 41 bapepiince in PG(t,) which con- tain G. Then Fn H; isa {on + Bow S00 z Pe vita} miner i in Hy for qe j +g +1 where the 6; are nonnegative integers such that > 6; = e. A Remark 1.2. (1) For any (t ~ 2)~flat @ in PG(t,q), there are q +1 hyperpl: PG(t,q) which contain G. a eae58 (2) There exists an {f,m;t,q}—minihyper F such that F C 9 for some 0—fat 0 in PG(t,q) if and only if there exists an {f,m0,q}—minibyper where 2 < 6 < and OSm 2, G)AZ%e=0, 41 =1, pa > 2andt> maton or) B22 mn 2 Rand t > ptm Let hy (i= 12-943) be nonnegative rs ‘, z integers such that 56 ethere exists a {evi + J yacs1s€00+ D> Yysit, q}—mininyper a a P such that (8) | FG [= Soeur for some (t ~ 2)-flat G in PG(t,q) and (>) FB; € Fo(Sj,41 ~ 1,42 —1)-++sn — itha) for any hyperplane Hy (1 < j $4 +1) which contains G, then F © Fiy(e,154125+++ baits )- Proposition L4, (Hamada (3). Let t, g, ¢, & and i (= 1,2,---yh) be any integers such that t > 2,q23,e=Oorl,1 2andt > mortentl, Pita fem + D mater + Erwatia} & ; & minihyper if and only if F € Fu(e, 1, 125--sHaits a). a a @)lnthe case A > Band t < pyr there is no fen + D tutewt aida} ; a a sminihyper. Proposition LS. (Hamada [7). Let t > 2,q>3and1 2p-+1, F is a {2ey41,20y5t,q}—minkbyper if and only if F € Fuswitsa). Definition L1. Let V be a flat in PG(t,q) where 2 <@ < t, A set $ of m points in V is called an m—arc in V if no 6+1 points in 5 are linearly dependent where m > 0 +1. For convenience sake, a set $ of @ points in V is also called an @~arc in V if @ points in are linearly independent. Remark 13. Let ex: and ex (j = 0,1,---y£—1) be nonnegative integers such that OStatenSe-L isa {Seam Sieymite}—mintype fo 5 = 1,2 and ot i A fs Fin Fy=0,then FU Ais {5 (ea tea) 5 6a eat} mie (), @ and (8) in Definition 11.2 ‘and (4) in Definition IZI.2), 87 Definition 12. Let V and W be # flat and a v—flat in PG(¢,q), respectively, such thot VW is an m—fiat in PO(t,q) where O Seiw for any (6 ~1)—fiat G in H, then [FH] > esvins- ist a Appendix II, Preliminary results in the case q=3 In this appendix, let ¢ = 3 and » = (3! — 1) /(3 ~1) for any integer 1 > 0. Proposition ILL. (Hamada and Helleseth [16], [17). Lett, a, and -y be integers such that either (@) 0 Sa=f<7 B+7+1, Fis {vasi + vpsi + Yyt1, Ya + Up + Yi t,3}— minibyper if and only if F € Fla,6,73t,3)- Definition ULL. (1) Let F(0,1,1;t,3) denote the family of all sets K in PG(é,$) such that K = V\$ for some 2-fiat V in PG(t,3) and some 4—are $ in V where t > 2 (cl. Definition £.1 in Appendix 1). (2 Let F(0,0,1;¢,3) denote the family of all sets K in PG(t,3) such that K = {(v1)s (0 + 1), (200+ 11), (02),("1 + ¥2),(b0% 211 +42)} for some noncollinear points (v0), (v1) and (v2) in PG(t,3). Remark ILL. In this appendix, (v) and (2v) represent the same point in PG(t,3) for any nonzero element v in the Galois field GF (3**). ‘Remark IL2. In (2) of Definition IT.1, let fo = 11, 1 = 2v9 + 2 and & = + ‘Then K = {(Go); (E1)s(E2)s (20 + &1)s (20 + &2)s (26a + €2)} (CE. (4) in Definition 17.2). Proposition 1L2. - (Hamada (6), [7). (1) In the case ¢ > 2 Fis a {2u; + v2,2v9 + v1;t,3}— minihyper if and only if either F € F(0,0, 3) or F € F(0,0,15t,3). (2) In the case t = 2, F is a {vy +2vs,m + 2v1;2,3}—minihyper if and only if F € F(0,1,1;2,3). @) In the case t > 3, F is a {01 + 202,00 + 20y;t,3}—minibyper if and only if either F € FQy1 143) oF F € F(0,1,154,8).58 Definition 112. (1) Let Fi(0,0,1,1;4,3) denote the family of all sets K in PG(t,3) such that = V \ $ for some 2~fiat V in PG(t,3) and some 3—arc S in V where t > 2. @ Let F3(0,0,1,154,8) denote the family of all sets K in PG(t,3) such that K = (V\ S$) U{P} for some 2—fiat V in PG(¢,3), some 4—arc $ in V and some point P ¢ V where ¢ > 3. @ La75(0,0,1,1; 4,3) deuore the tamlly of al sets Kin PG(t,3) such that K = LUK* for some 1—flat L in PG(t,3) and some minihyper K* in F(0,0,1;¢,3) such that LNK* = where 1 > 2. @) Let FA(0,0,1,1;¢,3) denote the family of all sets K in PG(t,3) such that K = {(o)s (61) Ea) (Es) (20 + €:) (2Eo +) (2k + Ea) (2E1 + €2), (241 + €3), (2a + &3)} for some four linearly independent points (£0), (€:), (G2) and (€3) in PG(t,3) where # > 3, Remark IL3. In Theorem 2.2 of Hamads, Helleseth and Ytrehus [18], let Go = 200, 4 = ei, 2 = eave and fs = cava. Thea the set of 10 points in Theorem 2.2 can be expressed as K in (4) of Definition 17.2. Proposition 11.3. (Hamada (6] and Hamada, Helleseth and Ytrebus [18). (1) Inthe case t= 2, F is a {201 + 202,209 + 201;2,3}— minihyper if and only if F € Fi(0,0,1,1;2,3). (2) Im the case t > 3, F is a {2m + 2v2,2vy + 201; ,3}—minihyper if and only if either PF € F(0,0,1,154,8) or F € F;(0,0,1,1;4,3) for some i in {1,2,3,4}. Appendix Il. Preliminary results in the case q=4 In this appendix, let ¢ = 4 and vy = (4! — 1)/(4—1) for any integer 1 > 0, __, Proposition IIL1. (Hamada and Helleseth [13)). Lett, a, f and 7 be integers such tht either (@) OS a B+7H, Pisa {vas + psi + Up4i, ta + ¥p + 2yit,4}— minihyper if and only if F € F(a,,7:1,4). aes Definition UL. Let F(0,0,1;t,4) denote the family of all ses K in PG(t,4) such that K = {(wo); (11); (12), (wo + wi), (wo + wa), (wi + w2), (wo + wi + wa)} for some ‘noncollinear points (wo), (wi) and (we) in PG(t,4) where ¢ > 2, Proposition HIL2. (Hamada (6). In the case ¢ > 2, F is a {20 + v2, 200 + m1jt,4}— ‘minibyper if and only if either F € ¥(0,0,1;t,4) or F € F(0,0,1;t,4). Definition IIL2, (1) Let Fi(0,0,1,1;4,4) denote the family of all sets K in PG(t,4) such that K = LoU Ly U{(eowa + wi +), (crm9 + aw + w2), (car + ony + ‘w2)} for ‘Some noncollinear points (we), (w1) and (w2) in PG(t,4) and some elements oo, e1 ald c2 59 in {0,1,a, a2} where t > 2, Lo = (wo) @ (wi), Li = (wo) @ (w2) and @ is a primitive clement in GF(2?) such that a? = a +1. @) Let Fx0,0,1,1}4,4) denote the family of all sets K in PG(t4) such that K = La U {on}, (os +02) (to + a1 + 2), (wo + aus + 12), (aa + ows +2), (wo + au + w2); (avo + aw; + wa)} for some noncollinear points (wo), (wi) and (wa) in PG(t,4) where 2 2 and Lo = (wo) © (wi). (3) Let Fil0,0,1,1;44) denote the family ofall seis K in PO(t,4) suck that K = (La\{(w1)}) U (L1\{(w2)}) U (La\{(ws + wa) { (ows + w2), (a? + w2)} for some noncollinear points (wo), (ws) and (w2) in PG(E,4) where # > 2, Lo = (wo) © (1), 21 = (w9) © (w2) and La = (wo) © (wi + wa)- . (4) Let Fal0, 0, 1,15#,4) denote the family of all sets K in PG(t,4) such that K = LUK* for some 1—flat Lin PG(t,4) and some minihyper K* in F(0, 0,1;t,4) such that LNK* = 0 where t > 3. Remark IIL1. If ¢2 = coa+c1a” in (1) of Definition I1T.2, then K in (1) contains three 1—flats Lo, Li and L* where L* = (cowo + wi + w2) ® (c1wo + aw; + w2) (Cf. Definition 1.2 with respect to the notation ©). Proposition 1113. (Hamada [5], (6). In the case ¢ = 2, F is a {201 + 2u2,200 + 20152,4}—minibyper if and only if F & (00,1, 1;2,4) for some é in 11.2.3}. Proposition I1L4, (Hamada, Helleseth and Ytrebus [18]. In the case t > 3, F is a (20) + 20,209 + 2vi;¢,4}—minihyper if and only if either F € F(0,0,1,1;44) ot F € F,{0,0,1,1;4,4) for some i in {1,2,3,4). Appendix IV. The correspondence between minihypers and codes meeting the Griesmer bound Let 5(k,q) be the set of all column vectors ¢, e'= (c0,¢1,-+-s¢4-1), in W(ke,q) such that either ey = 1 or e = 1, cig = cis 1 = 0, for some integer # in {0,1,---)k~2} where k > 3 and W(k,q) denotes a k-dimensional vector space consisting of column vectors over GF(q). Then 5(k,q) consists of (g* — 1)/(q—1) nonzero vectors in W(k,q) and there is no element « in GF(q) such that xa=ox for any two vectors x1 and x2 in 5(k,q). Hence the (g* — 1)/(q— 1) nonzero vectors in $(k,q) may be regarded. as (¢*—1)/(g—1) points in PG(E — 1,9). Proposition IV.1. (Hamada[4)). Let F be a set of f vectors in S(k,q) and let C' be the subspace of V(n,q) generated by a k x n matrix (denoted by G) whose column vectors are all the vectors in S(k, q) \ F where n = 4 — f,1< f 3 and 1 3 and 1 3 and 1 < d < gt, there is a one-to-one corres tween the set of all nonequivalent (n, kd; q]~codes meeting the Griesmer bound and the set of all {v4 — n, 4-1 — n+ d;k —1,g}—minihypers. Corollary IV.3. In the case k > 3 and d = gt! — e+ Seat) there is a one- to-one correspondence between the set of all nonequivalent [n, hyd; |~codes meeting the Griesmer bound and the set of all fens Bonnett South 144} ~minitypes where n = oy ~ (+ Soom): Corollary IV-4. In the case k > 3 and d = gi + Soe thre is « one-one correspondence between the st of all nonequivaleat fn, df odes meting he Grime ound and te set ofall { on, So oni — 19} pers where n= vy— So vate 61 References 1. AA. Bruen and R. Silverman, Arcs and blocking sets I Europ. J. Combin. 8 (1987), 351-356, 2. LH. Griesmer, A bound for error-cortecting codes, IBM J. Res. Develop. 4 (1960), 530-542. 3, N, Hamada, Characterization resp. nonexistence of certain gary linear codes attaining the Griesmer bound, Bull. Osaka Women’s Univ. 22 (1985), 1-47. N. Hamada, Characterization of minhypers in a finite projective geometry and its applications to error-correcting codes, Bull. Osaka Women’s Univ. 24 (1987), 1-24, 5, N. Hamada, Characterization of {12,2;2,4}—minchypers in a finite projective geometry PG(2,4), Bull. Osaka Women’s Univ. 24 (1987), 25-31. 6. N. Hamada, Characterization of {(q +1) + 2,1it,q}—minhypers and (2(g + 1) + 2,2;2,q}—min-hypers in a finite projective geometry, Graphs and Combin. 5 (1989), 6-81. 7. N. Hamada, A characterization of some (n, k,d;q)—codes meeting the Griesmer bound using a minihyper in a finite projective geometry, to appear in Discrete Math, In Chapter 4 in “Combinatorial Aspect of Design Experiments”, & N. Hamada and M. Deza, A characterization of some (n,k,d;q)—codes meeting the Griesmer bound for given integers k > 3, q > Sand d = gt) - g*—g — qt Sas B 3, q > 5) and its applications to error-comrecting codes, Discrete Math, 71 (1988), 219-231, 10. N. Hamada and M. Deza, A characterization of {p41 + ¢,%5%)q}—min-hyper and its applications to error-comecting codes and factorial designs, J. Statist, Plann. Inference 22, (1989), 323-336. 11,N. Hamada and M. Deza, A characterization of {2vaq1 + 20p41,20a + 20p;t,a}— minhypers in PG(t,9) (¢ > 2, > 5 and 0 3, q > 5, 1 <# < k—1) meeting the Griesmer bound, submitted for publication. 16,N. Hamada end T. Helleseth, A characterization of some {2va41 + ¥y+is2te+ vy;k — 1,3}—minihypers and some (n, k,3'-1 — 2-3 — 37;3)—codes (k > 3,0 < a'< 7 < k— 1) meeting the Griesmer bound, to appear in Discrete Math.62 17. N. Hamada and TT. Helleseth, A characterization of some minibypers in PG(t,3) and some temmary codes meeting the Griesmer bound, submitted for publication. 18 N. Hamada, T. Helleseth and @. Ytrehus, Characterization of {2(7 +1) + 2,2;t,q}— minibypers in PG(t,4) (¢ > 3, @ € {3,4}}), to appear in Discrete Math, 19, N, Hamada and F. Tamari, On a geometrical method of construction of maximal t—linearly independent sets, J. Combin. Theory 25 (A) (1978), 14-28. 20, N, Hamada and F. Tamari, Construction of optimal linear codes using flats and spreads in a finite projective geometry, Europ. J. Combin. 3 (1982), 129-141. 21, T. Helleseth, A characterization of codes meeting the Griesmer bound, Inform. and Control 50 (1981), 128-159, 22. R. Hill, Caps and codes, Discrete Math. 22 (1978), 111-137. 23, G. Solomon and 34. Suffer, Algebraically punctured cyclic codes, Inform. and Control 8 (1965), 170-179. 2A, F, Tamar, A note on the construction of optimal linear codes, J. Statist. Plann. Inference 5 (1981), 405411, 25. F. Tamari, On linear codes which attain the Solomon-Stifler bound, Discrete Math. 49 (1984), 179-191. Deligne-Lusztig Varieties and Group Codes by Johan P. Hansen? Matematisk Institut, Aarhus Universitet, Ny Munkegade, 8000 Aarhus C, Denmark ‘email: matjph@mi.aau.dk Abstract We construe algebraic geometric codes using the Deligne-Lusatig vatietica [De-Lu] asecated lo o connected reductive algebraic group @ defined over a Gnite fied Fy, wth Frobenius map P . The cole ate obtained a geometric Goppo codes, thats Her ronconrectng coda contre fom algerie vai [Gal] and (Go2) . The faite yroup GF of Lie type acla as Fy-rallonal a on tsi aa they” tome ‘mole ov ibe Beep ages F4[ OP]. Algae fcometic codes with a group algebra sirucare laduced from automorphisms of the underlying variety have been constructed and studied in [Hal] , [Ha2] , [Ha-St] and (vb “The Deligne-Lustig varieties used in the construction of the codes have in some cases many Fy-rational point, which ensures that the codes havea large word length. In case G is of type Ay the Del gneLuaig curve condeed have lee points over F 2. Incase G is a Suzuki group » Tespectively a Ree group 7G, Me Slee Can aves ediend Ee 1a? respectively 14@ points Geometric Gopps codes as group codes. Deligne-Lusitig varieties, Groupe of type 2, ‘Suzuki groupe #2, Ree groups 7G, pp. Weil's "explicit formulas” posewe ‘Supported by the Danish National Science Foundation,Introduction Many classical linear erro-correcting codes can be realised as ideals in group- algebras oF as modules over group-algebras.S. D. Berman [Be] elablished that the Reod-Muller codes over Fare ideals in a group-algebra over an elementary abelian 2-group. This was generalized by P. Charpia [Ch], [Ch2] ,(Ch3] and P. Landrock and O, Manz [La-Ma] , who showed that any Generalized Reed-Muller code is an ideal in a group-algebra over an elementary abelian p-group. In [Hal] , [Ha2] and [Ha-St] H. itenoth and the author construct algebraic geometric error-correcting codes with a group algebra structure. In [V] S- G. Viadut shows that the asymptotically good cover on classical and. Drineld ‘modular curves [T-V-Z] ,[M-V] constructed by Yu. I. Manin, M. A. Tsfasman, ‘Th. Zink and himself can be realized as group codes. ‘The general setup, which is treated in section 1, is to consider an algebraic variety X , defined over a finite field Fy, with a group G of F,-rational stomorphisms. On the variety X we consider an F,-rational and G-invariant divisor D together with a G-stable set P,,P,,....)Pn of Fy-tational points on X, none in the support of D. The associated Goppa code [Gol] and [Go2] is the image C= 4(L(D)) © FF of the Fy-linear map: @:L(D)' + FP fore (fC Pi) 25 S(Pa)) ‘The group G acts on the G-stable ‘set P,,Py,..-.4Pa_ of Fy-rational points giving F} a F,[G]-module structure. As the divisor ‘Dis G-stable, @ acts on L(D) and it becomes @ Fy{ G]-module. The Fy-linear map: o:L(D) =" FP becomes a F,[G]-morphism, and the geometric Goppa code C= 6(L(D)) © Fy ® F4{G]-module. In [Hal] a series of geometric Goppa codes is obtained from the Klein quartic, ‘codes which are ideals in the group-algebra Fg[G], where G is a Frobenius group of order 21. In (Ha2] , [St2] and [Ti] series of geometric Goppa codes are constructed from Hermitian curves, among these there are codes that are ideals in the sroup-algebra Fya{G] , G being @ non-abelian p-group of ordet q3, =p", P prime. In (1Ha~S¢] another series of group codes are presented. The codes are ideals in Fy{5] , where 5 is a Sylow-2-subgroup of order ¢? of the Suruki-group of order (a—1)(@-+1) and ¢= 2" ‘The codes are geometric Goppa codes over Fy with good parameters. Section 2 introduces another series of vari automorphisms. The varieties are Deligne-Luss reductive algebraic groups G defined over a finite field Fy, with Frobenius map :@— G [De-Lu] . Specifically, le G be a connected, reductive algebraic group G and let Xq be the Fe-scheme of all Borel subgroups of G with Frobenius morphism F:%q— Xg. For we W in the Weyl group X(w) © X, is the subsheeme of Xq of all Horel subgroups°B of G such that B and F() axe in elative posto wv. Let w= 4.-...74 be & minimal expen for w. X(syer50) is the space of sequences (By,...yBn) of Borel subgroups of G such that Bn FH and B,, and By are in relative position e or s;. The scheme 5 X(s,,---s4m) is of dimension n and it is a compactification of X(w). The Fy- tational points of X(ayy-ssyee)_ in X(e) and the finite group G of Lie type als as Fy-rational automorphisms on X(4,-..5n),X(w) and the Fy-rational pointe X(e) Section 3, treats the case where G is of type 742. Then the finite group GF has order @(q?—1 (+1) « For a simple reflection s¢ W in the Weyl sroup, the curve. (+) in izeducibe with genus =9) and it has 1+47+2gq F ,-rational points. This is the maximal number of Fe rational ts a curve of tht genus can have according to the Weil bound. Section 4 treats the case where Gis a Suruki-group. Then the finite group GF ata), = 2h, has onder” G(4—1)(Ge+1) = For a simple reflection s€ Win the Weyl group, the curve X() is irreducible with genus HoH fH 4 and it has 1449? Fy-rational points, This is the maximal number of Fy-rational points a curve ofthat genus can have according to Weil's explicit bound discused in the appendix. In [Tl-St} plane models have been given and resulting codes have been studied. Section 5 teats the case where G ia 9 Ree group, Then the finite group GF = 7G) gay has order P(E 1)(P +1). For & simple reflection #€ W in the Weyl group, the curve X(s) is irreducible with genus p= B= , (eso and it has 1+ 4° F,-rational points, which is the maximal number a curve of that genus can have according to Wells explicit bound discussed in the appendix. ‘The appendix discuss Weil's explicit formulas bounding the number of Fy rational points on the curves and thereby the length of the renlting geomet Goppa codes. I am grateful to J.P. Serre who suggested me to study Deligne-Lustig vatitiea in seach of curves with large groupe of automorphisms in relation to their genera. 1 Geomettic Goppa codes as group codes LI Let_X bea projective curve of genus g defined over finite field F Let D be a Fg-rational and positive divisor and let L(V) denote the Fg- vectorapace of rational functions defined over Fy such that f=0 of div()) B—D. Finally let P,,P,,----,Po_ be a set of F-fational points on X,, none in the support of D. The associated Goppa code {Gol] and [Go2] is the image C $(L(D)) © FP of the Fy-linear map: G14) $:(D) fo Ree Fe (Pn) «68 Theorem 1.2 (V. D. Goppa, ef [Gol}{Go2). Assume 0 < degD 29~2 (123) E> degDting ‘for degD < 29-2 Ins particular ti fedeieded. 1.3. From (1.24) is clear that geometric Goppa codes with good parameters aze to be found on curves where fis small that ison curves with a large number 1 of Fy-rational points compared to the genus g. As forthe number 1 ofall Fy- rational points on a curve, Weil's bound assert, thet (13.1) IW= (401 S29G With the “explicit formula” of Weil this general bound can in concrete cases be Improved. This technique is presented in the appendix and applied in the last 3 sections ofthis paper. 14 Lek @ be a group of Fy-rational automorphisms on the curve X ‘The action of G oa X induces an action on the divisors on X.. Assume that the divisor D is G-invariant and assume hat the set Py,Psy..-., Ps, of Fy-rational points on X is G-stable. The group G acts on the G-stable set P,,Pyyoo.n,Pa of F,-tational points giving F}a F4[G]-module structure. Av the dvieor Die G-invariant, G acts on L(D) by f= fog for fe L(D) and g@G y and it becomes let Fy{ G]-module. The Fy-linear map: $:(D) FP of (it) becomes a Fy[G]-morphism, and the geometric Gopps code $(L(D)) © FP 4 left F[G]-module. In case G acts freely and transitively on Py, Pay. then the’ Fy-vectorspoce on the points can be identified with the group algebra F,[G] and'the geometric Goppa code C= $(L(D)) © FP becomes a lef ideal in Fy[G]- 15 Let X be a projective curve of genus defined over a finite field of characteristic p . Let G denote the automorphism group of X. In case. p Hurwitz showed that G is finite and that (5.1) IG| < 84(9-1), p=0 Jn case_p>0 H. L. Schmidt showed that Gis inte, but |G] is not bounded as above. Stichtenoth [Sti] obtains (1.5.2) Io] < 64, P20 xcept inthe case where X ie defined by the afin equation (52.1) tye et, w>3 TL-W. Henn (He] obtains (15.3) Ig} < 80, p20 Gidea cae tas he ert tou can be sein o (15.3.1) IG] < 3(29)*{29 [claing 2 more cases. However in sotion we wil contact carves not on bi ist of excluded curves with more automorphisms than allowed by the proclaimed bound (15.3.1). or 2. Deligne-Lusatig varieties Let & be an algebraically closed field of characteristic p. 21 ‘The basic properties of affine algebraic groupe can be found in [Cal] and [C22] . Here we recollect what is needed for our purpose. An affine algebraic sroep G over bin an affine varity defined over which a alao « group such that the multiplication map Gx G-— @ and the inversion map @— Gare morphisms of varieties, Every alfine algebraic group is isomorphic to a closed subgroup of the ‘general linear group GLn(k) for some n. An affine algebraic group is called simple it has no non-trivial closed connected normal subgroup, 241.1 The multiplicative group &* ~ GL,(k) is an algebraic group. An algebeaie group isomorphic to E*x.-.x1* ls called a torus A Borel subgroup B ofa connected afine algebraic group’ G is a maximal connected salvable subgroup of G. Any two Borel subgroups of @ are conjugate in GA maximal torus lies in Some Borel subgroup of Gand two maximal tori in G are conjugate. ‘The group G has a maximal closed connected normal subgroup all of whee clementa are tnipotent. ‘This i the unipotent redical. The group Gis called reductive if the ‘nipotent radical is trivial: Let G be a connected, reductive algebraic groups defined over the fed . Let B be a Borel subgroup of G , let 7? be a maximal torus of G in B and let U be the unipotent radical of B. The Weyl group of G isthe finite group W = N(T)/T where N(T) isthe notinalizer Tin G- 24.2 Let Hom(‘,k*) be the character group of 7, that is the group of algebraic group homomorphiams from T’ to the multiplicative group &* and let likewise Y = Hom(&*,T) be the group of cocharacters of T. X and Y are fice abelian groups of the same finite rank. Let x€X and 7eY. By Composition we obtain & morphism, elrde so that x07 € Hom(é*,4*) and therefore ofthe form. (xe7)(Q) = 3 new for some integer m . ‘This gives a nondegenerate pairing XxY¥o7 Gor) 11> (xeq)(a) = AST vee ‘The groups X and Y are in duality and there i a bijection between themn xa ¥ where such that = 2, 2.1.3 Consider the finitely many minimal closed subgroups of U normalised by 7. Each of these are isomorphic to the additive group F. An element 1€ T° acts of these subgroups by conjugation, the corresponding automorhiam &— F is ‘multiplication by an element in £* . Hence the action of T’ by conjugation gives an clement in X= Hom(T,é*) for any minimal closed subgroupe of normalised by T. These elements are called positive roots. There is a unique Borel subgroup B- of G containing T such that BNB~=T. Let U~ be the unipotent radical of B=. As before we consider the minimal closed subgroups of U- normalised by T. The action of T by conjugation gives an element in X=68 Hom(T,4*) for any of these groups. These elements are called negative roots . ‘The action of the Weyl group W on T gives rise to actions of W on X and ¥ defined by OES) (wx) = x(wt(), (wy) d= w(7()), For each root @ there is an element wig € W such that, tig = woe and wh = 1 Let 0 CX be the set of roots. Then { wa | a€6)} generates W and such that the action of the Weyl group on X and Y" is determined by (2.1.3.2) a(x) = x~ g xex wal7) = 7 ~ a! rey 2.1.4 We have seen (2.1.3) that every connected reductive G has a es (2.1.3) ery group wewW, xeX, teT weW,yeY, ree (48, %,0¥) sssocated to it, where X,Y are the character and cocharacter groups of a maximal torus of G and @ is the set of roots and @Y the set of coroots. The root datum. (X,0,¥,0Y) uniguely determines the connected reductive group G. 2.1.8 Let G bea simple algebraic group. Let 6+ denote the positive roots (2.1.3). A positive root is called simple if it can not be expressed as the surm of two Positive roots. Let {a,...,a;} be the simple roots for G. The Cartan integers (2.15.1) Ay = takes on the value 2 if i=j and the values 0, i, —2, 3 if ij im such a way ‘that the integers @152) ay = Ay As takes on one of the values 0,1, 2,3. Let @153) 4 = Ma, € W fetal od let the order of 434; be my fen We ve A ( ¢ #i o ° 2 (2.1.5.4) 1 ° 3 2 2 4 3 s ‘ ‘The Dynkin diagram of G is a graph with I nodes corresponding to the simple Toots @;. The nodes corresponding to different simple roots a; and aj are connected by my = Aj Ay, bounds. An arrow is pointing from the ‘node corresponding to ‘a, to the node corresponding to a, if Ay; # —1. ‘The Dynkin diagrams interesting for our purpose are the following: Aa oo By — G sD 2.2 A connected reductive group G over & is isomorphic to a closed subgroup af GLq(t) for some m (2.1). Amap FryerG + cy is called o atendard Frotenins map if Frye isthe gestrition ofthe morphism Fro 4 on Glg(k) for some embedding of "G"ihto sorte GLn(k). A Frobenius map is a ™ F:G— 6 such that some power F is a standard Frobenius map. The finite groupe GF where G is a connected reductive group and F is a Frobenius map are called finite groups of Lie type. The real number Q defined by (2.2.0.1) Gar will be of later importance. ‘The fundamental theorem of Lang-Steinberg [LL] and [Ste] asserts that the where B= Frye mep (22.0.2) G+ G6, Us) =s'K(s) is surjective, This result has important consequences; in the following we recollect what is needed for our purpose. 221 Let @ be a connected reductive group, then G has « F-steble Borel subgroup. Namely let B be a fixed Borel subgroup, any other Borel subgroup is of the form gBg™ for some geG (2.1.1). The group F(B) is subgroup, #0 there is an element g€@ such that 9 F(B)(so (2.2.0.2) we find a g€G such that L(g) = 9" F(g) = go - Now gBg isa Festable Borel ‘subgroup as F(gBg) = F(g) F(B) P(g), (540) F(B)( 990)" = 98s. 22.2 Let G be a connected reductive group, then any two F-stable Borel subgroups are conjugate by an element in G"., For let By, B, be two F-stable Borel subgroups, then B, = 98g"! for some g@G (2.21). As both groups are F- stable we have 9"*P(9) € Ng(Bo)= By- By the theorem of Lang Steinberg applied tg, By there is a be By such that (4) = YF() = g'F(9), that is gh EG" and B, = (9b) Bo(go). 22.3 A Frobenius morphism induces @ graph automorphism p of the Dynkin diagram (2.1.5) when the arrows are disregarded. Let T be a maximally split torus. The Frobenius morphism induces an action on the character and cocharacter groups of T: Pix +X F(x) t = x(F()) xex and Pyay Fy) A = F(y(A)) vey ‘The action of F on the roots is related to the graph automorphism p of the Dynkin diagram. Specifically F(p(a)) is a positive multiple of a each positive root a. The Dynkin diagrams with F-action interesting for our purpose are the following twisted groups where the P-action permutes the two roots: 2, ~~ *By SS a ES Ss70 ‘2.2.4 1f G has type 7A, , the real number @ (2.2.0.1) can take any value 2, GEN, ef 0. These groupe are defined with reapect to © now te Hermitian form in 3 variables on F ga corresponding tothe involution A+ AY, 2.2.6 If G has type 7B, , then then characteristic p has to be 2 and the number Q (220.1) must satiety G@? = 2"*!, NEN. The finite groupe ‘= 7B,(Q?) ae the Suiuki groupe, 2.2.6 If G has type 7G, , then then characteristic p has to be 3 and the number Q (2.20.1) must satiety Q?= 3°"*?, wEN. ‘The finite groups = 7G,(Q") are the Ree groups. See eae ee ee ae rte rene tere Te Reset eoecanaNclaseeee ees Seale eer ed nearer cieminceae ee gelesen cel nee ener ener need ree antec ate stcior ee een arenas Saaremaa Set wane eae ae ae ea eee cal rade oneal Eee ee ie int aie aaa ene Pag Ei eH goad De 2 BEGET) oc ree teres i eee ena nc ate oteeaet arene oi orimee ea 232 The — subscheme X(w) © Xg (23.11) is smooth of (0) © Xg (23.1.1) is smooth of pure (321) dim X(w) = IC where w +4 is a minimal expresion for w as simple reflections [De-Lu, ir 233 Thembaheme X(v) © Xg (314) is OF otble Let = tyeety bea minimal expresin for Thea (aiseeeyon) space of sequences (py...,Bn) of Borel subgroups of @ such that Be FB, and Bj, and By are in relative position « ors, ‘The scheme 2 {fort l of dineson » and ti compaction of X(w) [De-Ly, 29-5 The Fyrrational points of X(1,,.-.,t) is Xe) and the finite sxoup Gof Lie type acts as F,-rational automorphisms on X( s,,---»4) » X(w) and the Fy-rational points X(e) 4m be a minimal expression for w as simple in) is irreducible if and only if any simple reflection SEW isin the F-orbit of some 5, where i=1,...,m [Lul,3:10d] 23,7 ‘The Buler characteristic of X(w) is according to [De-Lu, Theorem 7.1] determined by n o(G)—o(T) __1aFy 2.3.6.1 X(w)) = (=1 th G ) x(X(w)) = (-1) gl) where Tis a P-stable maximal torus contained in BEX(w), Stg is the Steinberg representation gf Grand o(G) (resp. o()) is the Fyrdik of G (cesp. T'). The order | 7" | is calculated by the formula (302) It] = ldeyog(eeF- 11, cf, [Ca2], where Yq = Hom(4*, Tq) be the group of cocharacters of Tp (2.21) » ‘where Ty is a F-stable maximal torus contained in a P-stable Borel subgroup and the action of F and w on Y is described in (2.2.3) and (2.1.3) . 24 In case s€ W is a simple reflection we obtain from (2.3) that the Deligne-Lusatig variety X(s) = X(s)UX(e) i a curve with the group GF of Lie type acting as Fé tomorphism. ‘The Fy-rational points on X(3) is X(e) and the curve X(s) is irreducible if and only if any simple reflection +¢ W isin the F-orbit of «. 24.1 ‘The genus and Buler characteristic of X(o) = X(s)UX(e) is determined by (2.4.11) 2— 29 = x(X(w))+x(X(e)) which is calculated using (2.3.61) and the number of Fy-rotional points on X(s) is x(X(e)) , which is also calculated using (2.3.6.1) . 3. Groupe of type 7A, — Hermite curves. There are two simple roots ay , ay « The corresponding Dynkin diagram has two nodes with 1 bound between them. ‘The Cartan matrix (2.1.5.1) ia (01) ‘The Frobenius morphism interchanges the two nodes. If G has type *A , the real number Q (2.2.0.1) can take any value p*, e€N, e#0. Let Q. ‘These groups are defined with respect to a non-degenerate Hermitian form z' + y'+2'= 0 ‘on F 2 corresponding to the involution X++ A". The finite groupe GP= 2A,(F) have order @(q?—1)(P +1). ‘The element vg, € W acts on the simple coroots according to (21.3.2) and the Cartan matrix (3.0.1) y Yo a¥ y way (gY) = gl - g¥ = pa} wat} eY— a! = af +'ay ‘The endomorphism of Yp@R induced by the element wa, € W therefore has matrix = [a3] ‘The endomorphism of Yp@R induced by the Frobenius has matrix(203) | a4 | ‘3.1 Now let Tbe a F-stable maximal torus contained in BE X(wa,) and let T, be a F-stable maximal torus contained in By€X(e) . Then according to (2.3.6.2) we obtain Gay InP] = dety, op(wayeF -1)| = -11 o4 ro etd ST-LE t]feecees 12) ITE] = Ldety, op leoP—1)] = =e-1 {i e}-[o st] PROPOSITION 3.2 Let G@ be @ connected reductive group of type "Ay aver BOE a. del wa, € W eo simple reflection and let X(q,) be the correspondily, Deligne-Lasitig variety. Then X(wq,) is en irreducible! curve of fa cH with 14 points over Fy. The finite group GP= 7Ag(q?) has onder ESSE a to rey FF so aaeamctee a) ‘The variety is a curve according to (2.3.2) and irreducible according to (2.36) as the Frobenius interchanges the two simple rools. From (2.3.7) we Aeiermine the Baler charctrsticn wing (11) and (312) B21 x(X(wa,)) = ("OC _1GFL_ ee, Sige) ITF] — SPH) _ P@=a) 7 = (F-10941) (8.22) x(X(e)) = (=1)"9)= 07a) _1 6 = Stg(e) [TI 3 #(P-1)P +1) ogee and the genus using 2.4.1.1 u (3.2.3) 2-2 = x(X(e)) + x(X(a,)) 148 (¢ -D(+1) ° isa ‘The finite group GP= 7Aq(q) order q°(q?—1)(¢ +1) and it acts ass group of Fj-tatonal automorphiama on X(wa,) by 23). PROPOSITION 3.3. The irreducible curve K(wa,) of genus pegs with 14+¢° points over F 4 has the mazimal number of rational points allowed by Be Wael ts fete ftn of bee teat! aE NO = GSD andthe number Nm of Fyu-rational points is determined by the formala Nn = 14 9"= (i 4(-O) 00 men ‘The claim follows from" the general theory of Zeta-functions (sce the appendix), as the formula for Nis seen to be true by inspection. REMARK 3.4 It is possible to construct geometric Goppa codes over Fg such that I distonce > 14g — PEL ‘The codes are modules over the group-ting F o[7Aq(q7)]. These codes have been studied in detail in (H2]- ¢ dimension + mi 4. Groups of type 7B, — Suzuki groupe. There are two simple roots ay , ‘1, . The corresponding Dynkin diagram has two nodes with 2 bounds between them and an arrow from the node corresponding to a, 10 that of a (2.2.4). The Cartan matrix (2.1.5.1) is 42-1 (40.1) aa ‘The Frobenius morphism is the two nodes, If @ has type 2B, , then peat : gn charatrticp bas to be 2 and the real number @ (2.201) must satiesty @= 2" nEN. Let g=Q. The finite groups G'= 7B,(¢) are the" ‘Suzuki groups and they have order q?(q—1)(4q?-+1) ‘The clement wq, € W acts on the simple coroots according to (2.1.3.2) and the Cartan matrix (4.0.1) ¥) =. 0¥ vs oY y wa,jeY) = yal — a¥ = -aY Fe A ‘The endomorphism of Y,@R induced by the element wa, € W therefore has matrix (402) [" : l ‘The endomorphism of Y@R_ induced by the Frobenius has matrix oA Stg is the Steinberg representation of 72,(¢) and assumes the value Stg(e) = 4.1 Now let_T’ be a F-stable maximal torus contained in BE X(wa,) and let T, be a F-stable maximal torus contained in B.€X(e) - Then according to (2.3.6.2) we obtain 141) ITP] = dety g(waeP—1)| = “11 o 8 0 fe seep [se] ee (a2) [TEL = Ldetyon(eoF- 01 = hs FE Ile PROPOSITION 42 Let G be o connected reductive group of type *By over k=Fq,q= 241, mEN. Let ua, € W be a simple reflection and let May) be the corresponding Deligne Laselig variety. Then X(Wy,) ie a irreducible curve of genus 1-54 with 144° points over Fy. The finite group 3B,(4) is a Suzuki group, it bas order’ @(q--1)(@P1) nd it acl coe gro of Fy. ralonal axlomorphions on X(ua,) ‘The variety is a curve according to (2.3.2) and irreducible according to (2.3.6) as the Frobenius interchanges the two simple roots. From (2.8.7) we determine the Euler characteristica using (4.1.1) and (4.1.2): = ete) _ 16 mary Sg IT -(¢+8(HP- 8 a-1) (421) x(X(wa,)) Par VPI) “FQ- Bath) o(G)—o(%) GF] 422) x(X(@) = (-1) Seg) ITF 2U-DE+)) _ Ofa-D gti and the genus using 24.1.1 (4.2.3) 2-28 = x(Xqcg,) + X(Xrca) = gei- (f+ B(GP-BH-1) co =f ‘The finite group GF= 7B,(4) is a Susuki group, it bas order @P'(¢—1)(q?-+1) and it acta as a group of Fg-rational automorphisms on X(ua,) by (23.3). We collect what we know about X(Wer,) in the following proposition. PROPOSITION 4.3 The irreducible curve X(wa,) of genus »-&24 with 149 points over Fy has the mex numberof rational points allowed by lhe caplicit formala” of Wei The Zeta fonction ofthe curve is C4 Bats at? “= 00-0) and the number Nm of F gn- rational points is determined by the formate A(XF (4) Nm = 1+e"—s96 4B if m=1,7 mod 8, 6=0 if m=2 mod if m=0 mod 8. where 6 4, b= if m=4 mod 8, 5= 5 mod 8 ond 6: ca ‘The claims follow from the determination ofthe eigenvalues of the Frobenius given in [Lu2] and the remarks in the appendix using $(0) = 1+ 2($eos(a) +4e00(20)) in (26) baving ols @= 434 , such thatthe a, of (a3) are gi (—¥ + f) imidiately giving the Zeta-function and consequentely the formulas fot Mm «76 REMARK 4A It is possible to construct geometric Goppa codes over Fy, 241 REN, such that dimension + minimal distance > L+q? — fH ‘The codes are modules over the group-ring Fg[7Aa(q)]- ‘These codes have been studied in detail in (Ha-St}. 5. Group codes from Ree groups. ‘There are two simple roots , ay . The ‘corresponding Dynkin diagram has two nodes with 3 bounds between them and an arrow from the node corresponding to a to that of a. The Cartan matrix is va [23] ‘The Frobenius morphism interchanges the two nodes. If @ has type 7G, , then then characteristic p has to be 3 and the rgal aumber @ (2.2.0.1) must satisfy Ga SEN. The finite groups P= 7G,(Q?) are the Ree groups. Let + then Faia). g= 94", has order P(q—1)( +1). clement tig, € W acta on the simple coroots according to (21.8.2) and the Cartan matrix (6.0.1) y y ‘ta, () =a} Wq,(a¥) = aY - aY = aY +3ay ‘The endomorphism of Y,@R induced by the element we, € W therefore has matrix (6.0.2) Tie endomorphism of Y40R ind by the Frobeie ht mt (603) ‘i [ oo | Sq the Sener erection of 3G,(4) and assumes the value Stg(e) = 5.1 Now let 7 be a F-stable maximal torus contained in B€ X(wa,) and let To be & F-stable maximal torus contained in Bee X(e) « Then according to (2.3.62) we obtain (eta) 18] = Lay, gg(sator = 101 = [Ee e} elf SF EE eee 7 (6.1.2) [TF] = Idety, galeoF— 11 = os 10 = Belles PROPOSITION 5.2 Let G be « connected reductive group of type 7a over E=F, 2M+1 NEN. Let wa, € W be @ simple reflection and let eum.) be the corresponding Delgne-Lusstig.verieiy, Then X(wq,) i a drvedatible curve of genus 7-8 @ , Woo with 144° points over Fy. The finite group GF= %G,(q) is « Ree group, it ci rt EN) and ite acts as 6 group of Fq-retonel ‘ulomorphisms on. (way). ‘The variety is a curve according to (2.3.2) and irreducible according to (2.36) as the Frobenius interchanges the two simple roots, From (27) we Aevermine the Euler characteristica using (6.1.1) and (6.1.2): e(G)—o(T) _1GPy Sig(e) | -(8 + 8 (grt e-4- BH-1) 24) 1) o(G)—o( To) es 622) xUinga) = (VO eset wi £G-DE+)) _ Cr and the genus using 2 = x(Xqcp,) + xrep) = et -( b+ BCG + ee - BH-1) e = R= w , (ed ‘The finite group GP= *G,(¢) is'a Ree group, it has order @°(¢—1)(° +1) and it acts as 2 group of Fg-rational automorphisms on X(wa,) by (23.8). We collect what we know about X(Wq,) inthe following proposition.8 PROPOSITION 5.3 The irreducible curve K(wa,) of genus 7-5 -M , “a9 with 144? points over Fy has the maxjpial number of rational points allowed by "explicit formulas” of Weil Let ao= 3 the Zeta function ofthe curve is aaa Marde+/2 14 dete 2) Da a2) tea ‘and the number Nm of F w-rational points is determined by the formula Nn = 144"= go4i(a—1)[(4-+3 40-41 )oosm/2+2(1 + 9) cosima/6_] ‘The claims follow from the determination of the eigenvalues of the Frobenius given in [Lu2] and the remarks in the appendix using 10) = 4(1 + $eos(a) +c08(20))? (a8) having roots 0= 4x counted with multiplicity « and) {z counted yb, such that the a, of (a3) are tif, (—¥ ih) G imidiately giving the ‘Zeta-function and consequentely the formula for Am - (cf. {PeD. REMARK 5.4 It is possible to construct geometric Gopps codes over Fy a= 31 aN, such that + minimal distance > ie (B® a 2) ‘The codes are modules over the group-ring Fg{*Gz(¢)]- dimensi REMARK 4.5 In [PE] curves with the same genera, Zetafunction and ‘automorphism groups have been realized as singular curves in P?- Appendix. Weil's "explicit formulas”. In [Sel] the "explicit formulas” of Weil for bounding the number of F g-rational points oa a curve of genus is treated. Let X be a curve defined over he finite field Fg . Let ay denote the number of primedivisors on X of degree d and let (@) Ym =D day, ml am be the number of Fym- rational points on X- ‘The Zetarfunction off X is the formal power series (2) Z%F M0) = e( Ym br), m2i which is a rational function of the form 79 Ta-ea-a,n) y 24M) = "Goya satiesfying the Riemann hypotheses oe a; = Wep(id;), o,eR ‘Taking the logarithmic derivative of Z(X,Fq)(#) we obtain = 1+" =( wm £ reee(ms) Consider the trigonometric expressions 5) y Deng 4, ded (6) 0) = 1+ 2 encase), HY) = ‘Then it follows from (a5) that DEH) +E bak) = 9 WI From this one concludes sbout the number N= N, of Fq- rational points on X. Proposition 0.8 Let the notation be as above and assume that f(@) > 0 for all GER and that cy 20 forall n> 1. Then veo, wy (CH) and eguality holds if and only if 4) ae (andes * ) fey 20 wd B bape = 0. ‘The Deligne-Lusstig curves associated to the algebraic groups of type 7A, 2g, and 7G, all have the maximal number Fg- rational points on X allowed by the above formulas as we have seen.[Be] [Cal] [Cs2] tout] {ot {oxs] [che] [De-La] [De-La} (601) [Go2] (wa) [Ha2) [Hest] (Hes) (He) (1) [12-Ma} (ut) [Lu2] [Pe] [Sel] S.D. Berman, On the Theory of Group Codes, Kibernetka, 3 (1967), 31-39. RW. Carter, Simple Groups of Lie Type, John Wiley and Sons Ltd, 1972. TW. Carter, Finite Groups of Lie Type, John Wiley aad Sons Ltd, 1985. P. Charpin, Codes idéaus de cgrlainesalgdbres modulaires, ‘These de Sime cycle, Universite de Paris VII (1982). P. Ghaspin, The extended Reed-Solomon codes considered as ideals of « modular eljebra, Annals of Discrete Math. 17 (1989), 171-176. P. Charpin, A new description of some polynomial Codes: the primitive generalized Reed-Muller codes, preprint. C. Chevalley, Introduction to the theory of algebraic functions of o sariable, New York 1951. L'Damgdrd and P. Landrock, [deals ond Codes in Group Algebra ‘Aathus Universitet, Preprint Sets. P. Deligne and G. Lustig, Representations of reductive groups over {finite field, Ann. of Math. 108 (1976), 103-161 YV. D. Goppa, Codes on algebraic Curves Dokl. Akad. NAUK, SSSR, 259 (1961), 1289-1200; (Soviet Math. Dok 24 (1981), 170-172.) V.D. Gopps, Aigebraice-geometric Codes, Inv. Akad NAUK, SSSR, 46 (1982); (Math. USS.R. levestiya, 21 (1983), 75.91) 43. P. Hansen, Codes on the Klein Quartic Ideas ond decoding, IEEE ‘Trans. on Inform. Theory 33 (1987). J. P. Hansen, Group Codes on Algebraic Curves, Mathematica Gotingensis, Heft 9, Feb. 1987. J.P. Hansen and. H, Stichtenoth, Group Godes on Certain Algebraic Curves with Mony Retional Points, Aathus Universite, Preprint Seria 1986/89 No. 7. (to. appear in Applicable Algebra in Engineering, ‘Communication and Computing). H. Have, Theorie der relat splischen algebraschen Funklionenkrper, 3. Reine Angew. Math. 172 (1954), 37-54. H. W. Henn, Funktionenkirper mil grosser Automorphismengrappe, 3. Reine u, Angew. Math, 302 (1978), 96-115. S. Lang, Algebraic Groups over finite Fields, Amer. J. Math. 78 (1958), 555-569. P. Landiock and 0. Manz, Classical Codes as Ideals in Group Algebras ‘Aatus Universite, Preprint Serco, 1986/87 No. 18, G. Lasatig, Representations of finile Chevelleygrowps Regional conference series in mathematics; no. 39. AMS 1978. G. Lusatig, Representations of finite Chevalley groups Re J.P. Pedersen, A Function Field related to the Ree group, to appeat in the proceedings. I.P, Serre, Sur le nombre des poins rationnels dune courbe alyébrigue ‘ur un corps fini, [se2] {se3] (st) [se2] {si3] [ste] im [t-V-2] {v]} at ©. R. Acad. Sci. Paris Sér. I Math, 296 (1983), 397-402. J.-P. Serre, Corps locaus, Patis 1962. J.-P. Serre, Annuaire du College de France, 1983-1984. H. Stichtenoth, On Aulomorphisms of Geometric Goppa Codes, (to appear in J. of Alg). H. Stichtenoth, A note on Hermitian Codes over GF(*), IEEE Trans. on Inform. Theory $4 (1988), 1345-1348. HL. Stichtenoth, Self dual Goppa Codes, 43. Pure Appl. Algebra 55 (1988), 199-211. R. Steinberg, Budomorphisms of linear algebraic groups, Mem. Amer. Math, Soc., 80 (1968). H. J. Tierama, Remarks on Codes from Hermitian Curves, IEEE Trans. on Inform. Theory 33 (1987), 605-609. M. A. Tafasman, S. G. Viadut and Th. Zink, Modular curves, Shimura curves and Goppa codes, better than Varshamov- Gilbert bound, Math. Nachr., 109 (1982), 13-28. 8. G. Viadut, Modular Codes as Group Codes, MAT-REPORT NO.1990- 25, Matematisk Institut, DanmarksTekniske Hojskole.Spectra of Linear Codes and Error Probability of Decoding Gregory L. KATSMAM, Michael A, TSFASMAN, Serge G. VLADUT (6, Profuesemaje at. Woecow 117342, U-5.5.R Introduction one of the most important characteristics of an error-correcting code is its spectrum. Therefore for asymptotic families of codes of growing length it is natural to consider the asymptotic behaviour of their spectra, Moreover the knowledge of this behaviour helps one to estimate the probability of error for maximum likelihood decoding for the considered family of codes. This paper is based on the study of spectra of algebraic-geonetric codes - and arbitrary linear codes as well - carried out in [1] ( cf. also [2] 1.1.3, 3.1.3). The main idea of that paper was to redevelop the enumerator Wo(xty) over the basis t(x-yy""J) and to. estimate the corresponding coefficients B, which have a lucid algebraic-geonetric interpretation. Let C bea linear (n,k,d),-code and a, be the number of codewords of weight i. The enumerator is given by woe = Bay! oe EBay. to poi 83 For a family F of codes with n—+>« we set ep(0) = Aim sup 2 205 Atyay « where cer, osws1, the limit is taken over all [n,k,d],-codes n—+«#, and [+] denotes the integer part. We would like to find out or at least to estimate the function ap(o) in terms of asymptotic paraneters of the family F, such as the rate R, the relative distance 6, the relative dual distanse 5*, etc. First (in Section 1) we use the estimate for 5, obtained in {2} and obvious relations between a, and B, to estimate 4, and ap(o) for arbitrary linear codes. Then (in Section 2) we give some subtler estimates for algebraic-geonetric codes. Bounds for ap(w) can be used to estimate the error probability of maximum likelihood decoding. Let x€C input of a q-ary symmetric channel with error probability per symbol p, let y be the output of maximum likelihood decoding. Then Pg(p) is defined as the probability for y to be different from x. For a family F we can set be an Bp(p) = Lim sup 3 log, Po(P) « then for the given rate R let E(,R) = Lim int Ep(P) the Limit being taken over all families F vhose rate equals R, i.e. £(p,R) corresponds to the error probability of decoding in the best family of codes with rate R. This is the function to estimate. The best estimate known is due to Gallager [3]. It is but natural that using algebraic-geometric codes and estimates for their spectra we can ameliorate the Gallager bound in many cases (Section 3).cy 2. Spectra of linear codes In this section we give sone upper bounds for the spectrum of an arbitrary linear q-ary code in terms of its parameters and its dual distance. Here we present sone asymptotic results (for sequences of codes of growing length). For the case of a fixed code of a "finite length" see Remark 3 below. We use the approach introduced in [1] (see also [2] 1.1.3). Let us recall some notation. Yet ¢ bea linear [n,k,d),-code, d* its dual distance. Its spectrun (A,) is defined as ay = al) = [ec] BI - of course 4,=1 and a,=0 for 1sfsd-1. te enunerator w,(xty) 1s defined as wot) =P E ayy). ich*¥) fa ue” We define B, from the decomposition aed - woGry) =a + EB, e-yy yt, 10 clearly ="E Payee. aT yea 0S We get the following obvious 4 ard Momma a ay nin BP) Zt is easy to check that ar ties 85 where KA ypeee dg) = dim Cpe) the subcode C(i,,++++4)) © being defined as the space of vectors having zeroes at the positions i,,...,4,, and the sum is taken over all subsets — (4,,.--,4,) © (1)-.-¢n) of, cardinality t, Lsi,s...si,sn. The following statement (see [2] thm. follows from the MacWillians identity. 1.2.26) easily Let isd*-1. Then = (et - a. Lemna 2. Together Lema 1 and Lana 2 show that ay te, Ge) - vetwa )(F) = G)CH] ve oe Lemma 3, A, + Gea. : ee eee eee FOL a) fs Lemma 4. a; = () tg }. vhere K(n- i, d) is the maximal possible dimension k of an In = i, k, d)_-code. 0 ) greode. Prooft B, wn AD ay a (9) BED, men we tn me st (BE) = CA et us establish an asymptotic upper bound for the spectrum. To do this the following notation looks appropriat Consider an [n,K,d)_-code ¢ with the dual distance d. tet R=R(C) =k/n, 8 =8(c) =an, Sta at(c) sd'yn, Rok, J=d, m let w= j/n and sectaglese) = 8209 4, consider a family of cotes F (cj) of groving Lensth. By an thuce of language. ve apesk then about "a codet of roving [igen ta onan enatder only fenition 7 mich that the Theft PeRG) sim R(G) dows exist. Note that any faniiy contains duck a subfunity. aot = 6(F) = Mm int 5(G)) + and s+ = 5/(F) = 1im ine 84(c,) . We would like to study the "Limit" value of eg(use) » Se 2 ap() = Lim sup 3 1095 Arnal + were 04041, the Limit is taken over all (nykyd]g-cod cer, and [1 denotes the integer part. aj2d ‘tor any “J, wo see that pio) «R. n—e, Since Here goes the principal result of this section. Let a3(u) = R= 1+ Hla) Hy being the q-ary entropy function, It is well known that, the spectrum of a random code asymptotically behaves as a5 (w) - In particular, the following statement is valid. Proposition 1, For any R there exists a family F of Linear codes such that R(F)=R and -* for a p(w) <= a gto) for 0 5.0 < 8G (R) bqy(R) S052, e sata - ug the inverse function to where Ogy(R) = HZ R) , HG) being che 1 functi ‘a the grary entropy. Ye styy (Reet + eHa-w) w= aul ‘ asst 87 ‘Theorem 1. RF) = Ry Let F be a family of linear q-ary codes vith 3(F) = 8, and 8(r) = 3+, Then -* for osu S7+ (see [4])+ ‘me bound of Theorem 2 is valid for any family of codes on algebraic curves provided the limit R(F) does exist. Using technique developed in [5] one can obtain a similar (but a better) bound for a “random code" on algebraic curves. We give here the answer only in the most important case of asymptotically maximal curves.92 let q=49 be an even power of a prime, let (x) bea family of asymptotically maximal curves, i.e. curves with aim HQ) a yg ea ge TF (the maximal possible ratio of the nuaber vf Fo-points to the genus). et 8, 3,, following conditions: + 8, €(0, (@=1)/q] be defined by the i. 0 <8, <8, <8,¢8,<¢(@-1/G5 ii, 3, and 4, are the roots of the equation ig) + Ly - 8) mates 441. 3, and 5, are the roots of the equation Hg(8) # (= a)loggig = 1) #2477 where yea 7= (VE - 1)". Then it is known (see [5], or [2] Theorem 3.4.11) that for any R there exists a choice of divisors D uch that the family F= (C= (1,?,D),) has the following relation between R and 8 (and in some precise sense almost any choice is such). If osss5, or 8,585 (G-a)/q then Rel-u(8)- If 6,4856, then R=1-7-5. 1£ 8,5858, or 3,58%5, then yew Fytmay = ea HEE) ay06) -2- 720. Yor a given R such that 1-a(8) #Rea—¥—8, oF Lrr-sjrRea— Hye) let ald)” be the pudigwe) solution of the last equation. Let 3,,(R) = H."(1 - R) , where Wz! is the inverse function to the q-ary entropy. ‘The following statement is proved in [4], the technique being the sane as in (51. Theoren 3. Let qx 49 be an even pover of a prime, let (1) be a family of asymptotically maxinal curves. Then for any R there exists a choice of divisors D such that the family F = ((,?,D)z) has the above relation betveen R and 8 and Tf Re lO, 1- Hy(S)] v (2 = Hy(s,), 1] then wo {0 for 0.5 W< 8Gy(R) La gle) for Sgy(R) sus. . te Re (1 ~ Hy(8), 2-8-7] v (1-8-7, 2 - HOB) then - 8 for 020 < ay (R) ° - #229) sus ag) #4} eGo) - Re NA- HEE) for ay sus 0, eu) for wsusa , vhere w,=1- (R+7)(q-1)/a- If Re (1-8,-4,1-8,-7) then = for gto) Osuci-R-7 R-vs0s1., ato) for 2 7 3. Error probability of decoding Let C bea linear [n, k, d]_-code. This code can be used to transmit information over a noisy gq-ary symmetric channel. This means that we transmit code words x¢C, and the error probability per symbol equals some p ; moreover the probability that a given symbol is distorted to another given symbol equals P/(q~ 1). The distorted word z is then transformed (decoded) into a code word y nearest to z (so called marimum likelihood or minimum distance decoding). Let P,(p) be the Probability of wrong decoding (y # x) .4 Tet F bea family of codes c. tet = lim sup 2 Ep(p) = Lim sup 5 los.Po(P) + and let E(p,R) = Lim int E,(p) over all families of codes whose rate equals R. Yet e(n,v.p) be the probability that for the received word z is nearer to a given y of veight ¥ than to x (note that ¢,(n,v.p) does not depend on y but enly on wv). The folloving result bounding P,(p) in terns of the spectrum is well known. xm oery Lemma 6. Po(p) = E Ayeg(n dP) + a r a1 Proof: If there is an error of decoding, y should be a code word of some weight i. The probability that y differs fron 0 is at most the sum of probabilities that y equals sone given non-zero code word. Let us calculate ¢,(n,1,p) and its asymptotical behaviour. Tet x, eqieP) = 1am Hogae,in, [un] ,p) « Lemma 7. a) €4(n,i,p) = n,n, ae Te penenen, ¢ ps o2 (Mai)n (6) (6323) @-2)""a-m) *e (& where the sum is taken over all triples (m,, my m,) of non-negative integers subject to the conditions msn-i, ntnsi, antnjed. D) eglsp) = 0 1ogq Bg(P) + vhere Ag(P) = Pla - 2)/(a ~ 2) + avPCBIZTET Proofs tet x=oe8, that the first n= 4 nan-d yerm, we(r) = 1. suppose positions are zeroes. Let zr" have non-zero entries in the first n-4 positions and 95, n,+n,44 non-zero entries in other positions. Besides we suppose that inn, positions the coordinates of z are equal to those of y- If m, m,, and m, are fixed then the probability of receiving 2 as an output of a q-ary symetric channel when the zero word vas transuitted equals (a) ay" (.) (24 (a-2)""(2-p) n,in,tn, a the distance between x and z equals n,+n,+n,, and between y and z it equals i-n,+n,. Thus e,(n/d,p) = ie) 2 (sf) cous” (4 ) (2s) ay a-p) and the first statement is proved. To prove the second statement we need to logarithmize the Fi 2, previous expression, and note that log is asymptotically small. So the asymptotic value of the sum equals the maximun of its summands. Tt is not difficult to see that this maximum is as stated in b) . Remark 6, It is clear that @,(p) for 0*p4s(q-1/d is an increasing concave function tangent to the vertical axis at p=0 and to the horizontal line 6@=2 at p=(q-U/a- We are ready to estimate the error probability for a given family F Leama 8. Ep(p) + max (ap(o) + W L0gg84(P)) ep (P) # max (ap: Pa Proof: We logarithmize the expression of Lemma 6 and note that asymptotically the sum equals the maximal summand. Then we use part b of Lenma 7.)The following statement is obtained by a direct calculation combining Lemma 8 and Proposition 1. ‘Theorem 4, Let (a ~ 2)8,(P) RP) = 1 - glee =aeSteT| * 044 Rip) Then Sqy(R)10gg64(P) for (R +a) + logy + (a - 1)84(P)) for R,(p) +R E(RsR) * Eg(P/R) = Remark 7. This bound is called the Gallager bound without expurgation. The Gallager bound with expurgation is E(P/R) * BG (PR) = = min (p(R1)+(p44)1 ayy) 740) ~p) 249) bapa PO )+(p+2) jog, ((P/(a-1)) (q-1)+(1-p) » which is better than ,(p,R) starting from some R, = R,(p) « Moreover, it is knovn that for R,+RsC=1-H,(p) we have the complete answer E(B,R) = EG(P.R) Because of that further on we are interested only in R 0 Note that in this case R

You might also like

• Her Body and Other Parties: Stories

  

  From Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  Rating: 4 out of 5 stars

  4/5 (822)
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  From Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  Rating: 4.5 out of 5 stars

  4.5/5 (122)
• A Man Called Ove: A Novel

  

  From Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  Rating: 4.5 out of 5 stars

  4.5/5 (4610)
• • Magazines
  • Podcasts
  • Sheet music
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  From Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  Rating: 3.5 out of 5 stars

  3.5/5 (231)
• Rise of ISIS: A Threat We Can't Ignore

  

  From Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  Rating: 3.5 out of 5 stars

  3.5/5 (137)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  From Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  Rating: 4.5 out of 5 stars

  4.5/5 (234)
• Sing, Unburied, Sing: A Novel

  

  From Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  Rating: 4 out of 5 stars

  4/5 (1104)
• On Fire: The (Burning) Case for a Green New Deal

  

  From Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  Rating: 4 out of 5 stars

  4/5 (74)
• The Perks of Being a Wallflower

  

  From Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  Rating: 4.5 out of 5 stars

  4.5/5 (2104)
• Fear: Trump in the White House

  

  From Everand

  Fear: Trump in the White House

  Bob Woodward

  Rating: 3.5 out of 5 stars

  3.5/5 (738)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  From Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  Rating: 4 out of 5 stars

  4/5 (98)
• Brooklyn: A Novel

  

  From Everand

  Brooklyn: A Novel

  Colm Tóibín

  Rating: 3.5 out of 5 stars

  3.5/5 (1947)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  From Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  Rating: 4.5 out of 5 stars

  4.5/5 (266)
• Yes Please

  

  From Everand

  Yes Please

  Amy Poehler

  Rating: 4 out of 5 stars

  4/5 (1898)
• Shoe Dog: A Memoir by the Creator of Nike

  

  From Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  Rating: 4.5 out of 5 stars

  4.5/5 (540)
• The Unwinding: An Inner History of the New America

  

  From Everand

  The Unwinding: An Inner History of the New America

  George Packer

  Rating: 4 out of 5 stars

  4/5 (45)
• Bad Feminist: Essays

  

  From Everand

  Bad Feminist: Essays

  Roxane Gay

  Rating: 4 out of 5 stars

  4/5 (1018)
• A Tree Grows in Brooklyn

  

  From Everand

  A Tree Grows in Brooklyn

  Betty Smith

  Rating: 4.5 out of 5 stars

  4.5/5 (1929)
• John Adams

  

  From Everand

  John Adams

  David McCullough

  Rating: 4.5 out of 5 stars

  4.5/5 (2409)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  From Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  Rating: 3.5 out of 5 stars

  3.5/5 (2259)
• Thompson R. J. 2011. GSFM - An Integrated Approach To Mine Haul Road Design PDF

  

  Document84 pages

  Thompson R. J. 2011. GSFM - An Integrated Approach To Mine Haul Road Design PDF

  Patman Mutombo

  100% (1)
• Bow Thruster Tb35t

  

  Document5 pages

  Bow Thruster Tb35t

  Giyan Arpero Mania

  No ratings yet
• Manhattan Beach: A Novel

  

  From Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  Rating: 3.5 out of 5 stars

  3.5/5 (792)
• Wolf Hall: A Novel

  

  From Everand

  Wolf Hall: A Novel

  Hilary Mantel

  Rating: 4 out of 5 stars

  4/5 (3811)
• The Emperor of All Maladies: A Biography of Cancer

  

  From Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  Rating: 4.5 out of 5 stars

  4.5/5 (271)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  From Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  Rating: 4 out of 5 stars

  4/5 (897)
• The Woman in Cabin 10

  

  From Everand

  The Woman in Cabin 10

  Ruth Ware

  Rating: 3.5 out of 5 stars

  3.5/5 (2521)
• Grit: The Power of Passion and Perseverance

  

  From Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  Rating: 4 out of 5 stars

  4/5 (590)
• The Outsider: A Novel

  

  From Everand

  The Outsider: A Novel

  Stephen King

  Rating: 4 out of 5 stars

  4/5 (1850)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  From Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  Rating: 4.5 out of 5 stars

  4.5/5 (474)
• The Light Between Oceans: A Novel

  

  From Everand

  The Light Between Oceans: A Novel

  M.L. Stedman

  Rating: 4.5 out of 5 stars

  4.5/5 (789)
• The Art of Racing in the Rain: A Novel

  

  From Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  Rating: 4 out of 5 stars

  4/5 (4203)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  From Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  Rating: 3.5 out of 5 stars

  3.5/5 (401)
• Call Center Vocabulary

  

  Document22 pages

  Call Center Vocabulary

  cgandara1990

  100% (3)
• The Glass Castle: A Memoir

  

  From Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  Rating: 4.5 out of 5 stars

  4.5/5 (1716)
• Principles: Life and Work

  

  From Everand

  Principles: Life and Work

  Ray Dalio

  Rating: 4 out of 5 stars

  4/5 (608)
• History of Modern Art Painting Sculpture Architecture

  

  Document7 pages

  History of Modern Art Painting Sculpture Architecture

  aer057

  0% (1)
• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  From Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  Rating: 4 out of 5 stars

  4/5 (5810)
• Angela's Ashes: A Memoir

  

  From Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  Rating: 4.5 out of 5 stars

  4.5/5 (441)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  From Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  Rating: 4.5 out of 5 stars

  4.5/5 (348)
• The Constant Gardener: A Novel

  

  From Everand

  The Constant Gardener: A Novel

  John le Carré

  Rating: 3.5 out of 5 stars

  3.5/5 (104)
• Little Women

  

  From Everand

  Little Women

  Louisa May Alcott

  Rating: 4 out of 5 stars

  4/5 (104)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  From Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  Rating: 4.5 out of 5 stars

  4.5/5 (844)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  From Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brené Brown

  Rating: 4 out of 5 stars

  4/5 (1092)
• Steve Jobs

  

  From Everand

  Steve Jobs

  Walter Isaacson

  Rating: 4.5 out of 5 stars

  4.5/5 (807)
• El Adjetivo

  

  Document6 pages

  El Adjetivo

  Nelfa Vargas Chahuaylacc

  No ratings yet
• Facing Codependence - Book Review

  

  Document1 page

  Facing Codependence - Book Review

  SPPrivate_

  0% (1)
• Infografia Motricidad

  

  Document2 pages

  Infografia Motricidad

  Gabriela Castillo

  No ratings yet
• Roy Thomas

  

  Document5 pages

  Roy Thomas

  Stephanie Valtierra

  No ratings yet
• Discurso Javier Sicilia

  

  Document3 pages

  Discurso Javier Sicilia

  Periódico Central

  No ratings yet
• Experiencia y Razón - Pastor

  

  Document11 pages

  Experiencia y Razón - Pastor

  Daniel Santiago SANCHEZ RINCON

  No ratings yet
• Top Executive MBA Colleges in India 2018 - Ranks, Fees, Cut-Off, Placement

  

  Document3 pages

  Top Executive MBA Colleges in India 2018 - Ranks, Fees, Cut-Off, Placement

  Ashish Singh

  No ratings yet
• La Restauración Del Muro y Las Puertas de Jerusalén

  

  Document4 pages

  La Restauración Del Muro y Las Puertas de Jerusalén

  Jose Bailon

  No ratings yet
• 3b Assessment of Working Conditions of Crop Agricultural Workers Aged 15 To 17 Phase II

  

  Document28 pages

  3b Assessment of Working Conditions of Crop Agricultural Workers Aged 15 To 17 Phase II

  Aviects Avie Jaro

  No ratings yet
• Concepción de Persona Desde El Punto de Vista Sociológico y Jurídico

  

  Document4 pages

  Concepción de Persona Desde El Punto de Vista Sociológico y Jurídico

  Raul Mejia Fiore

  90% (10)
• Pride and Prejudice: Art of Characterization

  

  Document3 pages

  Pride and Prejudice: Art of Characterization

  Rizwan Rao

  100% (2)
• Actividad # 7 Video

  

  Document10 pages

  Actividad # 7 Video

  Luz Dary Urrea Urrea

  No ratings yet
• Ricapa NL

  

  Document138 pages

  Ricapa NL

  Lizeth Lucchini Sanchez

  No ratings yet
• Qualitative Data Analysis - Summary of Merriam (2009) Book

  

  Document4 pages

  Qualitative Data Analysis - Summary of Merriam (2009) Book

  Safhira Auralia

  No ratings yet
• Divorcio

  

  Document4 pages

  Divorcio

  cecilia emperatriz hernandez

  No ratings yet
• Esquema Del Realismo Peruano

  

  Document1 page

  Esquema Del Realismo Peruano

  Juan Luis Gutierrez Urbano

  No ratings yet
• Singkamas Cultural Management

  

  Document1 page

  Singkamas Cultural Management

  Alex Matanog Jumamoy

  No ratings yet
• Certificado de Aportes Al Sistema de Seguridad Social y Parafiscales

  

  Document2 pages

  Certificado de Aportes Al Sistema de Seguridad Social y Parafiscales

  Andrés Forero

  No ratings yet
• Cuadro de Diferencias y Similitudes de La Tecnologia, Ciencia y Tecnica - U1 - T2

  

  Document3 pages

  Cuadro de Diferencias y Similitudes de La Tecnologia, Ciencia y Tecnica - U1 - T2

  ༒ᴅᴊ Joຮe TM ༒ Elvir

  No ratings yet
• Como La Perspectiva de Genero Cambia La Tecnica de Psicoanalisis Relacional

  

  Document21 pages

  Como La Perspectiva de Genero Cambia La Tecnica de Psicoanalisis Relacional

  Jessica Galeano

  No ratings yet
• 2 Types of PV Systems

  

  Document4 pages

  2 Types of PV Systems

  Isabel Guevara

  No ratings yet
• Educabilidad y Enseñabilidad

  

  Document2 pages

  Educabilidad y Enseñabilidad

  clara ramirez

  No ratings yet
• Victorias Milling Company

  

  Document2 pages

  Victorias Milling Company

  Anonymous ubixYA

  No ratings yet
• Boleh Di Download Jika Anda Mau Maju

  

  Document11 pages

  Boleh Di Download Jika Anda Mau Maju

  Salimuddin

  No ratings yet
• CA A 01 Introduzione

  

  Document56 pages

  CA A 01 Introduzione

  Mario Todisco

  No ratings yet
• Beautiful Raja Ampat

  

  Document11 pages

  Beautiful Raja Ampat

  Sherly Salsa Dewi

  No ratings yet