Professional Documents
Culture Documents
https://doi.org/10.1007/s40840-021-01101-2
Abstract
In this paper, we prove that if a nonconstant finite order meromorphic function f and
its n-th order difference operator nη f share a1 , a2 , a3 CM, where n is a positive
integer, η = 0 is a finite complex value, and a1 , a2 , a3 are three distinct finite complex
values, then f (z) = nη f (z) for each z ∈ C. The main results in this paper improve
and extend many known results concerning a conjecture posed by Chen and Yi in
2013.
Project supported in part by the NSF of Shandong Province, China (No. ZR2019MA029), the FRFCU
(No. 3016000841964007), the NSF of Shandong Province, China (No. ZR2014AM011) and the NSFC
(No. 11171184).
B Xiao-Min Li
lixiaomin@ouc.edu.cn
Yan Liu
1173670420@qq.com
Hong-Xun Yi
hxyi@sdu.edu.cn
123
X.-M. Li et al.
In recent years the difference variant of the Nevanlinna’s theory has been established
in [4,9] and, in particular, in [8], by Halburd–Korhonen and by Chiang-Feng, inde-
pendently. Using these theories, some mathematicians in the world began to consider
the uniqueness questions of meromorphic functions sharing values with their shifts,
and produced many fine works, for example, see [12,13,23]. In this paper, we will
consider a uniqueness question of finite order meromorphic functions that share three
values with their difference operators. This question is concerning a conjecture from
Chen–Yi [2].
For introducing the following results, we first give the definitions of the order, the
hyper-order and the exponent of the convergence of zeros of a nonconstant meromor-
phic function (see [3,14,20]):
Definition 1 For a nonconstant meromorphic function f , the order of f , the exponent
of the convergence of zeros of f , and the hyper-order of f , denoted as ρ( f ), λ( f )
and ρ2 ( f ), respectively, are defined as
logT (r , f ) logN r , 1f logn r , 1f
ρ( f ) = lim sup , λ( f ) = lim sup = lim sup
r →∞ logr r →∞ logr r →∞ logr
and
log logT (r , f )
ρ2 ( f ) = lim sup ,
r →∞ logr
respectively.
We recall the following results due to Heittokangas–Korhonen–Laine–Rieppo [13]
and Chen–Yi [2], respectively:
Theorem 1 ([13, Theorem 2.1]) Let f be a nonconstant meromorphic function of a
finite order, and let a1 , a2 , a3 be three distinct values in the extended complex plane.
If f (z) and f (z + η) share a1 , a2 , a3 CM, where η is a nonzero complex number, then
f (z) = f (z + η) for all z ∈ C.
Theorem 2 ([2, Theorem 1.2]) Let f be a transcendental meromorphic function such
that its order ρ( f ) is not equal to an integer or the infinity, let η = 0 be a finite complex
value such that η f ≡ 0, and let a and b be two distinct finite complex values. If
η f and f share a, b, ∞ CM, then 2 f (z) = f (z + η) for all z ∈ C.
Chen and Yi also gave the following example in [2]:
Example 1 ([2]) Let f 1 (z) = e z and f 2 (z) = e z+S(z) . Here, S(z) = d exp(τ z) +
a(τ −1)
τ , where d = 0 is a finite complex value and τ = log 2πi
2 . Then, we can verify
that S is a periodic function with a period equal to log 2. Moreover, we can verify that
ρ( f 1 ) = 1, ρ2 ( f 2 ) = ∞, and that f j (z + log 2) − f j (z) and f j (z) share 1, 2, ∞ CM
for 1 ≤ j ≤ 2.
123
Results on a Conjecture of Chen and Yi
Based upon Example 1, Chen and Yi posed the following conjecture in [2]:
Conjecture 1 ([2]) If the assumption “ρ( f ) is not equal to an integer or the infinity”
is omitted, then the conclusion of Theorem 2 still holds.
Regarding Conjecture 1, we recall the following results proved by Li–Yi–Kang [16]
and Cui–Chen [5], respectively, which studied Conjecture 1 based upon the assumption
of finite order:
Theorem 3 ([13, Theorem 1.1]) Let f be a nonconstant meromorphic function of
finite order, and let η = 0 be a finite complex value. If f and η f share a1 , a2 , a3
CM, where a1 , a2 , a3 are three distinct values in the extended complex plane, then
2 f (z) = f (z + η) for all z ∈ C.
Theorem 4 ([5, Theorem 1.1]). Let f be a nonconstant meromorphic function of finite
order, let η = 0 be a finite complex value, let n be a positive integer, and let a and
b be two distinct finite complex values . If nη f and f share a, b, ∞ CM, then
nη f (z) = f (z) for all z ∈ C.
Regarding Theorems 3 and 4, one may ask the following question:
Question 1 what can be said about the conclusion of Theorem 4, if we replace the
assumption “nη f and f share a, b, ∞ CM” with “nη f and f share three distinct
finite complex values a1 , a2 , a3 CM ” ?
We will prove the following result to resolve Question 1 completely.
Theorem 5 Let f be a nonconstant meromorphic function of finite order, let η = 0 be
a finite complex value, let n be a positive integer, and let a1 , a2 , a3 be three distinct
finite complex values. If nη f and f share a1 , a2 , a3 CM, then nη f (z) = f (z) for
all z ∈ C.
By Theorems 4 and 5 we get the following result:
Corollary 1 Let f be a nonconstant meromorphic function of finite order, let η = 0
be a finite complex value, let n be a positive integer, and let a1 , a2 , a3 be three
distinct values in the extended complex plane. If nη f and f share a1 , a2 , a3 CM,
then nη f (z) = f (z) for all z ∈ C.
Throughout this paper, by meromorphic functions we will always mean meromor-
phic functions in the complex plane. We adopt the standard notations of the Nevanlinna
theory of meromorphic functions as explained in (see [14,20,21]). It will be conve-
nient to let E denote any set of positive real numbers of finite linear measure, not
necessarily the same at each occurrence. For a nonconstant meromorphic function h,
we denote by T (r , h) the Nevanlinna characteristic of h and by S(r , h) any quan-
tity satisfying S(r , h) = o(T (r , h)), as r runs to infinity outside of a set of finite
logarithmic measure. Let f and g be two nonconstant meromorphic functions, and
let a be a value in the extended plane. We say that f and g share the value a CM,
provided that f and g have the same a-points in the complex plane, and each com-
mon a-point of f and g has the same multiplicities related to f and g. We say that
123
X.-M. Li et al.
f and g share the value a IM, provided that f and g have the same a-points in the
complex plane (cf. [20]). We say that b is a small function or a slowly moving tar-
get of f , if b is a meromorphic function satisfying T (r , b) = S(r , f ) (cf. [9,20].
We call f n 0 (z) f n 1 (z + η1 ) . . . f n k (z + ηk ) =: M(z, f ) a difference monomial in
f (z), the integer n 0 + n 1 + · · · + n k =: γ M is called the degree of the difference
monomial M(z, f ). If M1 (z, f ), M2 (z, f ), . . . , Mk (z, f ) denote difference mono-
l
mials in f (z), then a j (z)M j (z, f ) =: P(z, f ) is called a difference polynomial
j=1
in f (z) with degree γ P = max{γ M j : 1 ≤ j ≤ l} and coefficients a j such that
T (r , a j ) = S(r , f ) (see [9,15]).
In this paper, we also need the following definition of the difference operator of a
nonconstant meromorphic function:
2 Preliminaries
In this section, we will introduce some results used to prove the main result in the
present paper. First of all, we introduce the following result due to Gundersen–Tohge
[7]:
Lemma 1 ([7, Theorem 3]). If f and g are two nonconstant rational functions that
share two values CM and one value IM, then f = g.
Lemma 2 ([1]). Let f and g be two distinct nonconstant meromorphic functions that
share 0, 1, ∞ CM. If f is a Möbius transformation of g, then f and g assume one of
the following six relations: (i) f g = 1; (ii) ( f − 1)(g − 1) = 1; (iii) f + g = 1;
(iv) f = cg; (v) f − 1 = c(g − 1); (vi) ((c − 1) f + 1)((c − 1)g − c) = −c. Here, c
is a complex number satisfying c = 0, 1.
Lemma 3 ([8, Corollary 3.4]). Let f be a nonconstant finite order meromorphic solu-
tion of the difference equation P(z, f ) = 0, where P(z, f ) is difference polynomial in
f (z) and its shift f (z + η1 ), f (z + η2 ), . . . , f (z + ηk ), and let δ < 1. If P(z, a) ≡ 0
for a slowly moving target a, then
123
Results on a Conjecture of Chen and Yi
1 T (r , r + |η|)
m r, =o + o(T (r , f ))
f (z) − a rδ
for all r outside of a possible exceptional set with finite logarithmic measure, where
|η| = max1≤ j≤k {|η j |}. Moreover, the Nevanlinna deficiency satisfies
m r, 1
f −a
δ(a, f ) = lim inf .
r →∞ T (r , f )
Lemma 4 ([10, Lemma 8.3]) Let T : [0, +∞) −→ [0, +∞) be a nondecreasing
continuous function and let s ∈ R+ . If the hyper-order of T is strictly less than one,
i.e.,
log log T (r )
lim sup = ζ < 1,
r →∞ log r
and δ ∈ (0, 1 − ζ ), then
T (r )
T (r + s) = T (r ) + o ,
rδ
θn π θn π
Sj : − + (2 j − 1) + ε < θ < − + (2 j + 1) − ε,
n 2n n 2n
n+1
n+1
1
N r, +n N (r , f k ) < (μ + o(1))T (r , f j ), j = 1, 2, . . . , n,
fk
k=1 k=1
k= j
123
X.-M. Li et al.
Lemma 7 ([24, Proof of Theorems 1 and 2]). Let F and G be two distinct nonconstant
meromorphic functions that share 0, 1, ∞ CM. Suppose that N0 (r ) = S(r , f ). If F is
Möbius transformation of G, then N0 (r ) = T (r , F) + S(r , F). If F is not a Möbius
transformation of G, then N0 (r ) ≤ 21 T (r , F) + S(r , F), such that F and G assume
one of the following relations:
e(k+1)γ − 1 e−(k+1)γ − 1
(i) F = , G= ;
e −1
sγ e−sγ − 1
esγ − 1 e−sγ − 1
(ii) F = (k+1)γ , G = −(k+1)γ ;
e −1 e −1
esγ − 1 e−sγ − 1
(iii) F = −(k+1−s)γ , G = (k+1−s)γ .
e −1 e −1
Lemma 8 ([24, Lemma 8]). Let s and t be two relatively prime integers, and let c be
a finite complex number such that cs = 1, then there exists one and only one common
zero of χ s − 1 and χ s − c.
Lemma 9 ([22, Lemma 1]). Let f and g be two distinct nonconstant meromorphic
functions sharing 0, 1, ∞ CM. Then, there exist two entire functions α and β such
that
eα − 1 e−α − 1
f = β , g = −β ,
e −1 e −1
where eα ≡ 1, eβ ≡ 1, eβ−α ≡ 1. Moreover,
T (r , g) + T (r , eα ) + T (r , eβ ) = O(T (r , f )),
3 Proof of Theorem 5
Proof First of all, by Gundersen [6, Theorem 3] and the assumption that f and nη f
share a1 , a2 , a3 CM, we have T (r , nη f ) ≤ 3T (r , f ) + S(r , nη f ) and T (r , f ) ≤
3T (r , nη f )+S(r , f ). This implies that f is a nonconstant rational function if and only
123
Results on a Conjecture of Chen and Yi
f − a1 a2 − a3 nη f − a1 a2 − a3
F= · and G = n · . (1)
f − a3 a2 − a1 η f − a 3 a 2 − a 1
Then, by (1), the supposition f ≡ nη f and the assumption that f and nη f share
a1 , a2 , a3 CM, we deduce that F and G share 0, 1, ∞ CM such that F ≡ G. We
consider the following two cases:
Case 1. Suppose that F is a Möbius transformations of G. Then, it follows by Lemma
2.2 and the obtained result F ≡ G we know that one of the following five relations
can occur:
Anη f + B
f = , (2)
Cnη f + D
123
X.-M. Li et al.
By (5) and the supposition a1 = a2 we deduce that at least one of P1 (z, a1 ) = B−a1 =
0 and P1 (z, a2 ) = B − a2 = 0 holds, say
P1 (z, a1 ) = B − a1 = 0. (6)
for all r outside of a possible exceptional set with finite logarithmic measure. Here,
δ ∈ (0, 1) is a positive constant.
Since a1 is a Picard exceptional value of f , we deduce by (7) that
1 1 1
T (r , f (z)) = T r , = m r, + N r, + O(1)
f (z) − a1 f (z) − a1 f (z) − a1
T (r , r + |η|)
=o + o(T (r , f (z))) + O(1)
rδ
T (r , f (z))
=o + o(T (r , f (z))) + O(1) = o(T (r , f (z))) (8)
rδ
for all r outside of a possible exceptional set with finite logarithmic measure. This is
impossible.
Subcase 1.1.2. Suppose that (IV) holds. Then, 1 and c are two distinct Picard excep-
−a1
tional values of F. This together with (1) implies that a2 and c1ca13−1 are two distinct
2 −a1 )
Picard exceptional values of f , where c1 = c(aa2 −a3 . Next, by (3) and in the same
manner as in Subcase 1.1.1, we can get a contradiction.
Subcase 1.1.3. Suppose that (V) holds. Then, 0 and 1 − c are two distinct Picard
−a1
exceptional values of F. This together with (1) implies that a1 and c2ca23−1 are two
distinct Picard exceptional values of f , where c2 = (1−c)(a 2 −a1 )
a2 −a3 . Next, by (3) and in
the same manner as in Subcase 1.1.1, we can get a contradiction.
Subcase 1.1.4. Suppose that (VI) holds. Then, ∞ and 1−c 1
are two distinct Picard
−a1
exceptional values of F. This together with (1) implies that a3 and c3ca33−1 are two
a2 −a1
distinct Picard exceptional values of f , where c3 = (a2 −a3 )(1−c) . Next, by (3) and in
the same manner as in Subcase 1.1.1, we can get a contradiction.
Subcase 1.2. Suppose that C = 0. Then, it follows by (2) that 1/ f and Cnη f + D
share 0 CM. Combining this with the assumption that f and nη f share a1 , a2 , a3
CM, we deduce a j = −D/C for 1 ≤ j ≤ 3. Moreover, by AD − BC = 0 we deduce
|B| + |D| > 0. By rewriting (2) we have
123
Results on a Conjecture of Chen and Yi
Then, by (11) and Lemmas 3 and 4 we deduce (7) and (8), this is impossible.
Subcase 1.2.1.2. Suppose that (IV) holds. Then, 1 and c are two distinct Picard
−a1
exceptional values of F. This together with (1) implies that a2 and c1ca13−1 are two
2 −a1 )
distinct Picard exceptional values of f , where c1 = c(a
a2 −a3 . Next, by the definition of
P2 (z, f ) in (9) and in the same manner as in Subcase 1.1.1, we can get a contradiction.
Subcase 1.2.1.3. Suppose that (V) holds. Then, 0 and 1 − c are two distinct Picard
−a1
exceptional values of F. This together with (1) implies that a1 and c2ca23−1 are two
distinct Picard exceptional values of f , where c2 = (1−c)(a 2 −a1 )
a2 −a3 . Next, by the def-
inition of P2 (z, f ) in (9) and in the same manner as in Subcase 1.1.1, we can get a
contradiction.
Subcase 1.2.1.4. Suppose that (VI) holds. Then, ∞ and 1−c 1
are two distinct Picard
−a1
exceptional values of F. This together with (1) implies that a3 and c3ca33−1 are two
a2 −a1
distinct Picard exceptional values of f , where c3 = (a2 −a3 )(1−c) . Next, by the def-
inition of P2 (z, f ) in (9) and in the same manner as in Subcase 1.1.1, we can get a
contradiction.
Subcase 1.2.2. Suppose that D = 0. We consider the following four subcases:
Subcase 1.2.2.1. Suppose that one of (I), (II) and (III) holds. Then, it follows by (10)
that two of a1 , a2 and a3 are Picard exceptional values of f and nη f . Without loss
of generality, we suppose that a1 and a2 are Picard exceptional values of f and nη f .
Then, by noting that nη a = 0 for any finite value a, we have (4). By (4) and the
definition of P2 (z, f ) in (9) we have (10). By noting that D = 0 and a1 = a2 , we
deduce by (10) that at least one of P2 (z, a1 ) = 0 and P2 (z, a2 ) = 0 holds, say (11)
holds. Then, by (11) and Lemmas 3 and 4 we deduce (7) and (8), this is impossible.
Subcase 1.2.2.2. Suppose that (IV) holds. Then, 1 and c are two distinct Picard
−a1
exceptional values of F. This together with (1) implies that a2 and c1ca13−1 are two
2 −a1 )
distinct Picard exceptional values of f , where c1 = c(a
a2 −a3 . Next, by the definition of
P2 (z, f ) in (9) and in the same manner as in Subcase 1.1.1, we can get a contradiction.
Subcase 1.2.2.3. Suppose that (V) holds. Then, 0 and 1 − c are two distinct Picard
−a1
exceptional values of F. This together with (1) implies that a1 and c2ca23−1 are two
123
X.-M. Li et al.
e(k+1)γ − 1 e−(k+1)γ − 1
(VII) F = , G = ;
esγ − 1 e−sγ − 1
esγ − 1 e−sγ − 1
(VIII) F = (k+1)γ , G = −(k+1)γ ;
e −1 e −1
esγ − 1 e−sγ − 1
(IX) F = −(k+1−s)γ , G = (k+1−s)γ .
e −1 e −1
Here, γ is a nonconstant entire function, s and k(≥ 2) are positive integers satisfying
1 ≤ s ≤ k such that s and k + 1 are relatively prime. By (1), Lemma 8, the standard
Valiron–Mokhon’ko lemma (see [18]) and the expressions of F in (VII)–(IX) we
deduce
By (12) we deduce
where b7,n (= 0), b7,n−1 , . . . , b7,0 are finite complex constants, and n = ρ( f ) is a
positive integer. By (14) we deduce
n
γ (z + jη) = b7, j (z + jη) j = γ (z) + p7, j,n−1 (z) for 0 ≤ j ≤ n, (15)
j=0
123
Results on a Conjecture of Chen and Yi
Next we let b7,n = α7,n eiθ7,n , α7,n > 0 with θ7,n ∈ [0, 2π ). Moreover, for any
π
given positive number ε satisfying 0 < ε < 4n , we consider the following 2n angles:
θ7,n π θ7,n π
S7, j : − + (2 j − 1) +ε <θ <− + (2 j + 1) − ε, 0 ≤ j ≤ 2n − 1.
n 2n n 2n
(16)
and
By Remark 1 we have
n
n
nη f (z) = (−1)n− j f (z + jη). (19)
j
j=0
where
R1 eγ (z+ jη)
a2 (a3 − a1 ) + a1 (a2 − a3 )esγ (z+ jη) − a3 (a2 − a1 )e(k+1)γ (z+ jη)
= (21)
(a3 − a1 ) + (a2 − a3 )esγ (z+ jη) − (a2 − a1 )e(k+1)γ (z+ jη)
and
R2 e−γ (z)
a2 (a3 − a1 ) + a1 (a2 − a3 )e−sγ (z) − a3 (a2 − a1 )e−(k+1)γ (z)
= . (22)
(a3 − a1 ) + (a2 − a3 )e−sγ (z) − (a2 − a1 )e−(k+1)γ (z)
By (14) and Lemma 5 we can see that there exists a positive number R = R(ε) that
depends only upon ε, such that for |z| = r > R, Re{β7 (z)} > α7,n (1 − ε)r n sin(nε)
if z ∈ S7, j where j is even, while Re{β7 (z)} < −α7,n (1 − ε)r n sin(nε) if z ∈ S7, j
where j is odd. Combining this with (15), the assumption that a1 , a2 , a3 are three
123
X.-M. Li et al.
distinct finite complex values , the fact 1 ≤ s ≤ k and the fact that c is a finite complex
value satisfying c = 0, 1, we deduce for 0 ≤ j ≤ n that
lim R1 eγ (z+ jη) = a2 , lim R1 eγ (z+ jη) = a3 , (23)
|z|→+∞ |z|→+∞
z∈S7,1 z∈S7,0
lim R2 e−γ (z) = a3 and lim R2 e−γ (z) = a2 . (24)
|z|→+∞ |z|→+∞
z∈S7,1 z∈S7,0
By (20)–(24) we deduce
n
n
n n
0 = a2 (−1)n− j = lim (−1)n− j R1 eγ (z+ jη)
j |z|→+∞ j
j=0 z∈S7,1 j=0
= lim R2 eγ (z+ jη) = a3 (25)
|z|→+∞
z∈S7,1
and
n
n
n n
0 = a3 (−1)n− j = lim (−1)n− j R2 (γ (z + jη))
j |z|→+∞ j
j=0 z∈S7,0 j=0
= lim R2 eγ (z+ jη) = a2 . (26)
|z|→+∞
z∈S7,0
and
By Remark 1 we have (19). By substituting (27) into (19), and then into (28) we
deduce
n
n
(−1)n− j R3 eγ (z+ jη)
j
j=0
123
Results on a Conjecture of Chen and Yi
with
R3 eγ (z+ jη)
a2 (a3 − a1 ) + a1 (a2 − a3 )e(k+1)γ (z+ jη) − a3 (a2 − a1 )esγ (z+ jη)
= (30)
(a3 − a1 ) + (a2 − a3 )e(k+1)γ (z+ jη) − (a2 − a1 )esγ (z+ jη)
n
n
(−1)n− j R3 eγ (z+ jη) = R4 e−γ (z) . (32)
j
j=0
and deduce
lim R4 e−γ (z) = a1 and lim R4 e−γ (z) = a2 . (34)
|z|→+∞ |z|→+∞
z∈S7,1 z∈S7,0
By (32)–(34) we deduce
n
n
n n
0 = a2 (−1)n− j = lim (−1)n− j R3 eγ (z+ jη)
j |z|→+∞ j
j=0 z∈S7,1 j=0
= lim R4 e−γ (z) = a1 (35)
|z|→+∞
z∈S7,1
and
n
n
n n
0 = a1 (−1)n− j = lim (−1)n− j R3 eγ (z+ jη)
j |z|→+∞ j
j=0 z∈S7,0 j=0
= lim R4 e−γ (z) = a2 . (36)
|z|→+∞
z∈S7,0
123
X.-M. Li et al.
and
By Remark 1 we have (19). By substituting (37) into (19), and then into (38) we
deduce
n
n
(−1)n− j R5 eγ (z+ jη) = R6 eγ (z) . (39)
j
j=0
where
R6 eγ (z)
a2 (a3 − a1 ) + a1 (a2 − a3 )e(k+1−s)γ (z) − a3 (a2 − a1 )e−sγ (z)
= (40)
(a3 − a1 ) + (a2 − a3 )e(k+1−s)γ (z) − (a2 − a1 )e−sγ (z)
and
R5 eγ (z+ jη)
a2 (a3 − a1 ) + a1 (a2 − a3 )e−(k+1−s)γ (z+ jη) − a3 (a2 − a1 )esγ (z+ jη)
= (41)
(a3 − a1 ) + (a2 − a3 )e−(k+1−s)γ (z+ jη) − (a2 − a1 )esγ (z+ jη)
and that
lim R6 eγ (z) = a3 and lim R6 eγ (z) = a1 . (43)
|z|→+∞ |z|→+∞
z∈S7,1 z∈S7,0
123
Results on a Conjecture of Chen and Yi
n
n
n n
0 = a3 (−1) n− j
= lim (−1)n− j R5 eγ (z+ jη)
j |z|→+∞ j
j=0 z∈S7,0 j=0
= lim R6 eγ (z) = a1 (44)
|z|→+∞
z∈S7,0
and
n
n
n n
0 = a1 (−1)n− j = lim (−1)n− j R5 eγ (z+ jη)
j |z|→+∞ j
j=0 z∈S7,1 j=0
= lim R6 eγ (z) = a3 . (45)
|z|→+∞
z∈S7,1
eβ̂1 − 1 e−β̂1 − 1
F= , G= , (47)
eβ̂2 − 1 e−β̂2 − 1
where eβ̂1 , eβ̂2 and eβ̂2 −β̂1 are nonconstant entire functions. Moreover,
By [6, Theorem 3] and the obtained result that F and G share 0, 1, ∞ CM, we
have
123
X.-M. Li et al.
By (49) and the obtained result that eβ̂1 and eβ̂2 are nonconstant entire functions we
see that α and β are nonconstant polynomials. Therefore,
and
where b̂k,n k , b̂k,n k −1 , . . . , b̂k,0 are finite complex constants with b̂k,n k = 0 for 1 ≤
k ≤ 2, while n k is a positive integer for 1 ≤ k ≤ 2.
By (52) and (53) we deduce
n1
β̂1 (z + jη) = b̂1, j (z + jη)k = β̂1 (z) + p̂1, j,n 1 −1 (z) (54)
k=0
and
n2
β̂2 (z + jη) = b̂2, j (z + jη)k = β̂2 (z) + p̂2, j,n 2 −1 (z) (55)
k=0
and
By Remark 1 we have (19). By substituting (56) into (19), and then into (57), we
deduce
n
n
(−1)n− j R̂1 eβ̂1 (z+ jη) , eβ̂2 (z+ jη) = R̂2 e−β̂1 (z) , e−β̂2 (z) , (58)
j
j=0
123
Results on a Conjecture of Chen and Yi
where
R̂1 eβ̂1 (z+ jη) , eβ̂2 (z+ jη)
and
a (a − a ) + a (a − a )e−β̂2 − a (a − a )e−β̂1
R̂2 e−β̂1 , e−β̂2 =
2 3 1 1 2 3 3 2 1
. (60)
(a3 − a1 ) + (a2 − a3 )e 2 − (a2 − a1 )e β̂1
−β̂ −
iθ
Next we let b̂ j,n j = α̂ j,n j e j,n j , α j,n j > 0 with θ j,n j ∈ [0, 2π ) for 1 ≤ j ≤ 2.
Moreover, for any given positive number ε satisfying 0 < ε < 4nπ j , we consider the
following 2n j angles for 1 ≤ j ≤ 2 :
θ j,n j π θ j,n j π
Ŝ j,k : − + (2k − 1) +ε <θ <− + (2k + 1) − ε, (61)
nj 2n j nj 2n j
where 0 ≤ k ≤ 2n j − 1 for 1 ≤ j ≤ 2.
We consider the following two subcases.
Subcase 2.2.1. Suppose that
and
a (a − a )
lim R̂2 e−β̂1 (z) , e−β̂2 (z) =
2 3 1
= a2 . (66)
|z|→+∞ (a3 − a1 )
z∈ Ŝ1,0
123
X.-M. Li et al.
n
n
n n
0 = a2 (−1) n− j
= lim (−1)n− j R̂1 eβ̂1 (z+ jη) , eβ̂2 (z+ jη)
j |z|→+∞ j
j=0 z∈ Ŝ1,0 j=0
= lim R̂2 e−β̂1 (z) , e−β̂2 (z) = a2 . (67)
|z|→+∞
z∈ Ŝ1,0
By (67) we have a2 = 0. Combining this with the assumption that a1 , a2 , a3 are three
distinct finite complex values, we have a1 a3 = 0. Therefore, it follows by (59) and
(60) that (58) can be rewritten as
n
n eβ̂1 (z+ jη) − eβ̂2 (z+ jη)
(−1)n− j
j (a3 − a1 ) + a1 eβ̂1 (z+ jη) − a3 eβ̂2 (z+ jη)
j=0
n
By multiplying (a3 − a1 ) + a1 eβ̂1 (z+kη) − a3 eβ̂2 (z+kη) on two sides of (68) we
k=0
have
n
n
(−1)n− j eβ̂1 (z+ jη) − eβ̂2 (z+ jη) H1, j (z)
j
j=0
n
= e−β̂1 (z) − e−β̂2 (z) (a3 − a1 ) + a1 eβ̂1 (z+kη) − a3 eβ̂2 (z+kη) , (69)
k=1
where
n
H1, j (z) = (a3 − a1 ) + a1 eβ̂1 (z+kη) − a3 eβ̂2 (z+kη) . (70)
k=0
k= j
n
n
(−1)n− j eβ̂1 (z+ jη) − eβ̂2 (z+ jη) H2, j (z)
j
j=0
n
= eβ̂2 (z)−β̂1 (z) − 1 (a3 − a1 ) + a1 eβ̂1 (z+kη) − a3 eβ̂2 (z+kη) , (71)
k=1
123
Results on a Conjecture of Chen and Yi
where
n
H2, j (z) = eβ̂2 (z) (a3 − a1 ) + a1 eβ̂1 (z+kη) − a3 eβ̂2 (z+kη) . (72)
k=0
k= j
n
n
(−1)n− j eβ̂1 (z+ jη) − eβ̂2 (z+ jη) H2, j (z)
j
j=0
n
n
= (a3 − a1 )n eβ̂2 (z) (−1)n− j eβ̂1 (z+ jη) − eβ̂2 (z+ jη)
j
j=0
+ Pn+2 eβ̂1 (z) , eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη) , (73)
n
eβ̂2 (z)−β̂1 (z) − 1 (a3 − a1 ) + a1 eβ̂1 (z+kη) − a3 eβ̂2 (z+kη)
k=1
= eβ̂2 (z)−β̂1 (z) − 1 Pn eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
+ (a3 − a1 )n eβ̂2 (z)−β̂1 (z) − 1 , (74)
where Pn+2 eβ̂1 (z) , eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη) is a poly-
n
n
nomial in eβ̂1 (z+ jη) and eβ̂2 (z+ jη) with nonzero constants as the coeffi-
j=0 j=0
cients and with degree equal to n + 2 such that the degree of each monomial of
Pn+2 eβ̂1 (z) , eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
n
n
is a polynomial in eβ̂1 (z+ jη) and eβ̂2 (z+ jη) with nonzero constants as the
j=1 j=1
coefficients and with degree equal to n such that the degree of each monomial of
Pn eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
123
X.-M. Li et al.
is greater than or equal to 1. Therefore, it follows by (73) and (74) that (71) can be
rewritten as
n
n β̂2 (z) n
(a3 − a1 ) e (−1)n− j eβ̂1 (z+ jη) − eβ̂2 (z+ jη)
j
j=0
+ Pn+2 eβ̂1 (z) , eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
− eβ̂2 (z)−β̂1 (z) − 1 Pn eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
= (a3 − a1 )n eβ̂2 (z)−β̂1 (z) − 1 . (75)
On the other hand, by (52)–(55) and the left equality of (64) we deduce
for 0 ≤ k ≤ n. Here, β̂1,k,n 1 −1 (z) and β̂2,k,n 1 −1 (z) are constants or nonconstant
polynomials of degree less than or equal to n 1 − 1. Therefore, by (76) we have
n1
= Ak (z)eb̂1,n1 z , (78)
β̂1 (z+kη) b̂1,n 1 z n 1 β̂1,k,n 1 −1 (z)
e =e e
and eβ̂2 (z+kη) = eb̂1,n1 z eβ̂2,k,n1 −1 (z)
n1
(79)
with
and
T r , eβ̂2 (z)−β̂1 (z) − 1 + T (r , Ak (z)) + T r , eβ̂1,k,n1 −1 (z)
+ T r , eβ̂2,k,n1 −1 (z) = o(T (r , eβ̂ j (z+kη) )) for 0 ≤ k ≤ n. (82)
123
Results on a Conjecture of Chen and Yi
By (77)–(82) and the obtained result that the degree of each monomial in
Pn+2 eβ̂1 (z) , eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
is greater than or equal to 3, and the obtained result that the degree of each monomial
in
Pn eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
is greater than or equal to 1, we deduce that eβ̂2 −β̂1 − 1, eβ̂1,k,n1 −1 , eβ̂2,k,n1 −1 , Ak are
small entire functions of each monomial of
Pn+2 eβ̂1 (z) , eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη)
and
Pn eβ̂1 (z+η) , . . . , eβ̂2 (z+nη) , eβ̂2 (z+η) , . . . , eβ̂2 (z+nη) for 0 ≤ k ≤ n.
By (83), the assumption a1 = a3 and the obtained result that eβ̂2 −β̂1 − 1 is not a
constant, we get a contradiction.
Subcase 2.2.1.2. Suppose that n 1 = n 2 and |b̂1,n 1 | < |b̂2,n 2 |. Then, by (52), (53),
(61)–(63) we deduce
and so it follows by (52)–(55), (58)–(63), (84), Lemma 5 and Hayman [11, p.7] that
n
a2 (a3 − a1 ) n
(−1)n− j
a3 − a1 j
j=0
n
n
= lim (−1)n− j R̂1 eβ̂1 (z+ jη) , eβ̂2 (z+ jη)
|z|→+∞ j
z∈ Ŝ1,1 j=0
= lim R̂2 e−β̂1 (z) , e−β̂2 (z) = lim R̂3 eβ̂2 (z)−β̂1 (z) , eβ̂2 (z)
|z|→+∞ |z|→+∞
z∈ Ŝ1,1 z∈ Ŝ1,1
a1 (a2 − a3 )
= = 0. (85)
a2 − a3
123
X.-M. Li et al.
Here,
and so it follows by (52)–(55), (58)–(63), (87), Lemma 5 and Hayman [11, p.7] that
n
a2 (a3 − a1 ) n
(−1)n− j
a3 − a1 j
j=0
n
n
= lim (−1)n− j R̂1 eβ̂1 (z+ jη) , eβ̂2 (z+ jη)
|z|→+∞ j
z∈ Ŝ1,1 j=0
= lim R̂2 e−β̂1 (z) , e−β̂2 (z) = lim R̂4 eβ̂1 (z)−β̂2 (z) , eβ̂1 (z) = a3 = 0, (88)
|z|→+∞ |z|→+∞
z∈ Ŝ1,1 z∈ Ŝ1,1
where
123
Results on a Conjecture of Chen and Yi
Subcase 2.2.2.2. Suppose that n 1 = n 2 and that there exists some nonnegative odd
integer k1 satisfying 1 ≤ k0 ≤ 2n 1 − 1 such that
lim eβ̂1 (z)−β̂2 (z) = ∞, lim eβ̂1 (z) = ∞, lim e−β̂2 (z) = ∞, (94)
|z|→+∞ |z|→+∞ |z|→+∞
z∈ Ŝ1,0 z∈ Ŝ1,0 z∈ Ŝ1,0
and so it follows by (52)–(55), (58)–(61), (89), (92)–(94), Lemma 5 and Hayman [11,
p. 7] that
n
−a3 (a2 − a1 ) n
(−1)n− j
−(a2 − a1 ) j
j=0
n
n
= lim (−1)n− j R̂1 eβ̂1 (z+ jη) , eβ̂2 (z+ jη)
|z|→+∞ j
z∈ Ŝ1,0 j=0
= lim R̂2 e−β̂1 (z) , e−β̂2 (z) = lim R̂4 eβ̂1 (z)−β̂2 (z) , eβ̂1 (z)
|z|→+∞ |z|→+∞
z∈ Ŝ1,0 z∈ Ŝ1,0
a1 (a2 − a3 )
= = 0. (95)
a2 − a3
where ϑ1 and ϑ2 are two real numbers satisfying 0 ≤ ϑ1 < ϑ2 ≤ 2π, and that φ
denotes the empty set. By (52), (53), (61), (97) and Lemma 5 we deduce
123
X.-M. Li et al.
and so it follows by (52)–(55), (58)–(61), (89), (98), Lemma 5 and Hayman [11, p. 7]
we deduce
n
−a3 (a2 − a1 ) n
(−1)n− j
−(a2 − a1 ) j
j=0
n
n
= lim (−1)n− j R̂1 eβ̂1 (z+ jη) , eβ̂2 (z+ jη)
|z|→+∞ j
z∈ Ŝ1,0 ∩ Ŝ2,l1 j=0
= lim R̂2 e−β̂1 (z) , e−β̂2 (z) = lim R̂4 eβ̂1 −β̂2 , eβ̂1 = a1 = 0.
|z|→+∞ |z|→+∞
z∈ Ŝ1,0 ∩ Ŝ2,l1 z∈ Ŝ1,0 ∩ Ŝ2,l1
(99)
Acknowledgements The authors wish to express their thanks to the referee for his/her valuable suggestions
and comments.
References
1. Al-khaladi, A.H.H.: Meromorphic functions that share three values with one share value for their
derivatives. J. Math. (Wuhan) 20(2), 156–160 (2000)
2. Chen, Z.X., Yi, H.X.: On sharing values of meromorphic functions and their differences. Results Math.
63(1–2), 557–565 (2013)
3. Chen, Z.X.: Complex Differences and Difference Equations. Science Press, Beijing (2014)
4. Chiang, Y.M., Feng, S.J.: On the Nevanlinna characteristic of f (z + η) and difference equations in the
complex plane. Ramanujan J. 16(1), 105–129 (2008)
5. Cui, N., Chen, Z.X.: The conjecture on unity of meromorphic functions concerning their differences.
J. Differ. Equ. Appl. 22(10), 1–20 (2016)
6. Gundersen, G.G.: Meromorphic functions that share three or four values. J. Lond. Math. Soc. 20(3),
457–466 (1979)
7. Gundersen, G.G., Tohge, K.: Unique range sets for polynomials or rational functions. In: Progress in
Analysis, pp. 235–246 (2003)
8. Halburd, R.G., Korhonen, R.J.: Nevanlinna theory for the difference operator. Ann. Acad. Sci. Fenn.
Math. 31(2), 463–478 (2006)
9. Halburd, R.G., Korhonen, R.J.: Difference analogue of the lemma on the logarithmic derivative with
applications to difference equations. J. Math. Anal. Appl. 314(2), 477–487 (2006)
10. Halburd, R.G., Korhonen, R.J., Tohge, K.: Holomorphic curves with shift-invariant hyperplane preim-
ages. Trans. Am. Math. Soc. 366(8), 4267–4298 (2014)
123
Results on a Conjecture of Chen and Yi
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.
123