Discussiones Mathematicae

General Algebra and Applications 40 (2020) 37–48

doi:10.7151/dmgaa.1327

ANALYTIC PROPERTIES OF THE APOSTOL-VU

MULTIPLE FIBONACCI ZETA FUNCTIONS

Utkal Keshari Dutta and Prasanta Kumar Ray

Department of Mathematics, Sambalpur University

Jyoti Vihar, Burla-768019, India
e-mail: utkaldutta@gmail.com
prasantamath@suniv.ac.in

Abstract

In this note we study the analytic continuation of the Apostol-Vu multi-

ple Fibonacci zeta functions
X 1
ζAV F,k (s1 , . . . , sk ; sk+1 ) = s1 s2 sk sk+1 ,
1≤m1 <···<mk
Fm 1 Fm2 · · · Fm k Fm1 +m2 +···+mk

where s1 , . . . , sk+1 are complex variables and Fn is the n-th Fibonacci num-
ber. We find a complete list of poles and their corresponding residues.
Keywords: analytic continuation, Fibonacci numbers, multiple Fibonacci
zeta function, Apostol-Vu multiple Fibonacci zeta function.
2010 Mathematics Subject Classification: 11B37, 11B39, 30D30, 11M99.

1. Introduction

The Euler-Zagier’s multiple zeta function ζEZ,k of depth k is defined by

X 1
(1.1) ζEZ,k (s1 , . . . , sk ) = ,
m1 s 1 m2 s 2 · · · mk s k
1≤m1 <m2 ···<mk <∞

where s1 , . . . , sk are complex variables [10]. Zhao [11] gave an analytic continu-
ation of (1.1) as a function of si (i = 1, . . . , k) to Ck using Gelfand and Shilov’s
generalized functions. Mehta et al. [6] showed the meromorphic continuation
of multiple zeta functions (1.1) with their poles and residues by means of an
38 U.K. Dutta and P.K. Ray

elementary and simple translation formula for specified multiple zeta functions.
The series
∞ X
X 1
(1.2) ,
m=1 n<m
ms1 ns2 (m + n)

was first introduced by Apostol and Vu [1] and they obtained the partial results
on its analytic continuation. The meromorphic continuation of (1.2) and more
general series
∞ X
X 1
(1.3) s1 ns2 (m + n)s3
,
m=1 n<m
m

to the whole space was first proved in [3]. In [4], Matsumoto generalized (1.3)
and introduced Apostol-Vu r-ple zeta function

ζAV,r (s1 , . . . , sr ; sr+1 )

(1.4)
X 1
= .
ms11 ms22 · · · msrr (m1 + m2 + · · · + mr )sr+1
1≤m1 <m2 <···<mr <∞

He proved that the series (1.4) is convergent absolutely when Re(si ) > 1 (1 ≤
i ≤ r), Re(sr+1 ) > 0 and can be continued meromorphically to the whole space
Cr+1 .
Let {Fn }n≥0 be the sequence of Fibonacci numbers and is recursively defined
by
F0 = 0, F1 = 1, Fn+1 = Fn + Fn−1 , for n ≥ 1.
√
αn −β n 1+ 5
The closed form expression for {Fn }n≥0 is Fn = α−β , where α = 2 and
√
1− 5
β = 2 are the zeros of Fibonacci characteristic equation
P∞ −−s x − 1 = 0. In x2
[7], Navas presented Fibonacci Dirichlet series ζF (s) = m=1 Fm , Re(s) > 0
for s ∈ C, where Fm denotes the m-th Fibonacci number and obtained that
ζF (s) is analytically continued to a meromorphic function on the complex
P∞ plane
C. Rout and Panda [9] considered balancing zeta function ζB (s) = m=1 Bm −s ,

Re(s) > 0 for s ∈ C, where Bm denotes the m-th balancing number and gave that
ζB (s) is meromorphically continued all over C. They also obtained that ζB (−m)
is an irrational number when m is an odd natural number. In a P consequent
paper, Behera et al. [2] proved the analytical continuation of ζC (s) = ∞ −s
m=1 Cm ,
Re(s) > 0 for s ∈ C, where Cm denotes the m-th Lucas-balancing number and
ζC (−m) is a rational number for any odd natural number m. In [8], Rout and
Meher defined the multiple Fibonacci zeta function of depth k as:
X 1
(1.5) ζF (s1 , . . . , sk ) = s1 s2 sk .
Fm 1 Fm2 · · · Fmk
1≤m1 <m2 <···<mk <∞
Analytic properties of the Apostol-Vu multiple Fibonacci ... 39

They derived the analytic continuation of ζF (s1 , . . . , sk ) of depth 2 and found

a complete list of poles and their corresponding residues. Recently, Meher and
Rout [5] proved the meromorphic continuation of multiple Lucas zeta functions
of depth d :
X 1
ζU (s1 , . . . , sd ) = ,
Un1 · · · Unsdd
s1
0<n1 <n2 ···<nd

where (Un ) is the Lucas sequence of first kind.

In this note, we present Apostol-Vu multiple Fibonacci zeta function and
explore its analytic continuation. We also estimate the residues corresponding to
their respective poles.

2. Analytic continuation of Apostol-Vu multiple Fibonacci zeta

functions

In this section, the Apostol-Vu multiple Fibonacci zeta function is introduced

and its analytic continuation is examined.
Definition 1. We define Apostol-Vu multiple Fibonacci zeta functions ζAV F,k as
X 1
(2.1) ζAV F,k (s1 , . . . , sk ; sk+1 ) = s1 s2 sk sk+1 ,
F F · · · Fmk Fm1 +m2 +···+mk
1≤m <···<m <∞ m1 m2
1 k

where s1 , . . . , sk+1 are complex variables.

The sum s1 + · · · + sk+1 is called the weight of ζAV F,k (s1 , . . . , sk ; sk+1 ) and
k is called its depth. One can observe that ζAV F,k (s1 , . . . , sk ; 0) is the multiple
Fibonacci zeta function (1.5).
Proposition 2. The series (2.1) is absolutely convergent in
 Xk 
k+1 
Dk+1 = (s1 , . . . , sk ; sk+1 ) ∈ C  Re(sj )+(k+1−d)Re(sk+1 ) > 0, 1 ≤ d ≤ k .
j=d

Proof. Let sj = σj + itj ∈ C and σj = Re(sj ) > 0, j = 1, 2, . . . , k + 1. From

(2.1), we can write
X 1
s1 s2 sk sk+1
1≤m <···<m
F F
m1 m2 · · · Fmk Fm1 +m2 +···+mk
1 k
∞ ∞
X 1 X 1
(2.2) = s1 s1 ···
m1 =1
Fm1 m =1 Fm1 +m2
2
∞ ∞
X 1 X 1
s sk s .
F k−1
mk−1 =1 m1 +m2 +···+mk−1 mk =1
Fm F k+1
1 +m2 +···+mk km1 +(k−1)m2 +···+mk
40 U.K. Dutta and P.K. Ray

Now,
∞   ∞  
X  1 X 1
 = (α − β)σ1
  
 αm1 − β m1
s1 
 s  
 Fm1 
1
m1 =1 m =1 1
∞
X 1
(2.3) ≤ (α − β)σ1
s1
m1 =1 α
m1 σ1 1 − (| αβ |)m1
 σ1 X ∞
α−β 1 1
≤ σ1 m1
= Λσ1 (α − β)σ1 σ
1 − |β/α| m1 =1
α α 1 − 1

and similarly,
 
 s 1  ≤ Λσ (α − β)σd 1
 
(2.4) F d d

, 2 ≤ d ≤ k,
ασd m1 +···+md

m1 +···+md

and
 
 1  ≤ Λσ (α − β)σk+1
 1
(2.5)  s
 F k+1 k+1

,
ασk+1 km1 +(k−1)m2 +···+mk

km1 +(k−1)m2 +···+mk

1
where Λσj = (1−|β/α|)σj
. By virtue of (2.2), (2.3), (2.4) and (2.5),
 
X  1 
sk+1
 
s1
 Fm sk
1 · · · Fmk Fm1 +···+mk

1≤m1 <···<mk
∞   ∞  
 1  X
 s1
X  
≤  s 
 Fm1  F 1
 ···

m1 =1 1
m2 =1 m1 +m2
∞
 ∞ 
  X  
 s 1 1
X  
sk+1
  
 F k−1   F sk F 
m k−1=1 m 1 +···+m k−1 m =1 k
m 1 +···+m k km1 +(k−1)m2 +···+mk

∞ ∞
σ1 σk+1
X 1 X 1
≤ Λσ1 Λσ2 · · · Λσk Λσk+1 (α − β) · · · (α − β) σ2 (m1 +m2 )
m1 =1
ασ1 m1 m2 =1
α
∞ ∞
X 1 X 1 1
×···×
mk−1 =1
ασk−1 (m1 +m2 +···+mk−1 ) mk =1
ασk (m1 +···+mk ) ασk+1 (km1 +(k−1)m2 +···+mk )
Pk+1 ∞ ∞
σj
X 1 X 1
= Λσ (α − β) j=1

m1 =1
α(σ1 +···+σk +kσk+1 )m1 m2 =1
α(σ2 +···+σk +(k−1)σk+1 )m2

∞
X 1
×···×
mk =1
α(σk +σk+1 )mk
Analytic properties of the Apostol-Vu multiple Fibonacci ... 41

Pk+1
σ
j=1 j 1 1
= Λσ (α − β) 2
ασ1 +···+σk +kσk+1 − 1 ασ2 +···+σk +(k−1)σk+1 − 1

1
×··· ×
< ∞,
ασk +σk+1 −1

as α > 1, where Λσ = Λσ1 . . . Λσk Λσk+1 and Λσ1 , . . . , Λσk and Λσk+1 are the
positive constants depending on σ1 , . . . , σk and σk+1 respectively. Therefore, the
series (2.1) converges absolutely in the domain Dk+1 . This completes the proof.

Theorem 3. The Apostol-Vu multiple Fibonacci zeta function ζAV F,k (s1 , . . . , sk ;
sk+1 ) can be analytically continued to a meromorphic function on Ck+1 . It has
possible simple poles on the hyperplanes

sd + · · · + sk + (k + 1 − d)sk+1 = −2(rd + · · · + rk + (k + 1 − d)rk+1 )

πi rd + · · · + rk + (k + 1 − d)rk+1 + 2a
+ ,
log α
1 ≤ d ≤ k,

with r1 , . . . , rk+1 ∈ Z≥o and a ∈ Z.

Proof. For any s = σ + it ∈ C, we have
α − βm s
 m s
√ −s ms
 m  
s β
Fm = √ = ( 5) α 1−
5 α
∞    mk
− 2s ms
X s k β
(2.6) = 5 α (−1)
k α
k=0
∞
X s 


− 2s m+1 k m s−2k
= 5 (−1) α .
k
k=0

As α > 1, the series (2.6) converges. Using the above identity in (2.2), we get

ζAV F,k (s1 , . . . , sk ; sk+1 )

∞ ∞ ∞
X X X 1 1 1 1
= ··· s1 s2 · · · sk sk+1
Fm1 Fm1 +m2 Fm1 +m2 +···+mk Fkm +(k−1)m +···+m
m1 =1 m2 =1 mk =1 1 2 k
∞ ∞  
X s1 X −s1
= 52 (−1)(m1 +1)r1 α−m1 (s1 +2r1 )
r1
m1 ,m2 ,...,mk =1 r1 =0
∞  
s2 X −s2
×52 (−1)(m1 +m2 +1)r2 α−(m1 +m2 )(s2 +2r2 ) × · · ·
r2
r2 =0
42 U.K. Dutta and P.K. Ray

∞   P Pk
−sk
sk



1+ kd=1 md rk − md (sk +2rk )
X
×5 2 (−1) α d=1
rk
rk =0
∞   Pk
−sk+1
sk+1


(−1) 1+ d=1 (k+1−d)md rk+1
X
(2.7) ×5 2
rk+1 =0
rk+1
Pk

− (k+1−d)md (sk+1 +2rk+1 )
×α d=1 .

−s   
i  ri −|si |
Using the fact   ≤ (−1) for 1 ≤ i ≤ k + 1, we have

ri ri
∞ ∞ ∞  
X X X  1 1 1 1 
··· 
 Fm s1 s2 · · · sk sk+1

m1 =1 m2 =1 mk =1 1 F m 1 +m 2
Fm 1 +m 2 +···+m k
Fkm 1 +(k−1)m 2 +···+m k

∞  ∞  
X  s1 X −s1
= 5 2 (−1)(m1 +1)r1 α−m1 (s1 +2r1 )
m1 ,m2 ,...,mk =1

r1 =0
r 1
∞  
s2 X −s2
×52 (−1)(m1 +m2 +1)r2 α−(m1 +m2 )(s2 +2r2 ) × · · ·
r2 =0
r 2
∞   Pk Pk
−sk
sk X



×52 (−1) 1+ d=1 md rk α− d=1 md (sk +2rk )
rk =0
rk
∞   Pk Pk

−sk+1
sk+1 X



1+ (k+1−d)m r − (k+1−d)m (s +2r )

×5 2 (−1) d=1 d k+1
α d=1 d k+1 k+1 
rk+1 =0
rk+1

∞ Pk+1 ∞  
X n=1 σn −m1 σ1
X −|s1 |
≤ 5 2 α (−1)r1 α−2m1 r1
m1 ,m2 ,...,mk =1 r1 =0
r1
∞  
X −|s2 |
× α−(m1 +m2 )σ2 (−1)r2 α−2(m1 +m2 )r2 × · · ·
r2 =0
r2

Pk ∞   Pk
−|sk |

X

−
×α d=1 md σk
(−1)rk α−2 d=1 md rk
rk =0
rk
Pk ∞   Pk
−|sk+1 |



× α d=1 (k+1−d)md σk+1 (−1)rk+1 α−2 d=1 (k+1−d)md rk+1
X

r =0
rk+1
k+1
∞ Pk+1 −|s1 |
σ 
n=1 n
X
= 5 2 α−m1 σ1 1 − α−2m1 α−(m1 +m2 )σ2
m1 ,m2 ,...,mk =1
 −|s2 | Pk
 Pk
−|sk |
× 1 − α−2(m1 +m2 ) × ···× α d=1 md σk
1 − α−2 d=1 md
Analytic properties of the Apostol-Vu multiple Fibonacci ... 43

Pk
 Pk
−|s
k+1 |
×α d=1 (k+1−d)md σk+1
1 − α−2 d=1 (k+1−d)md

∞ Pk+1 −|s1 | −|s2 |

σ
n=1 n
X  
≤ 5 2 α−m1 σ1 1 − α−2 α−(m1 +m2 )σ2 1 − α−2
m1 ,m2 ,...,mk =1
Pk

−|sk | Pk
 
md σk d=1 (k+1−d)md σk+1
× · · ·× α d=1 1−α−2 ×α 1−α−2 −|sk+1 |

Pk+1 Pk+1 ∞

X
σ
n=1 n −2 − |sn |
α−(σ1 +···+σk +kσk+1 )m1


= 5 2 1−α n=1

m1 =1
∞
X ∞
X
α−(σ2 +···+σk +(k−1)σk+1 )m2 · · · α−(σk +σk+1 )mk < ∞.
m2 =1 mk =1

Thus, the above series absolutely converges for a fixed point (s1 , . . . , sk ; sk+1 ) in
Dk+1 . Then, by interchanging the order of summation in (2.7), we get

ζAV F,k (s1 , . . . , sk ; sk+1 )

Pk+1 ∞   ∞   ∞  
n=1 sn
X −s1 X −s2 X −sk+1
= 5 2 (−1)r1 (−1)r2 · · · (−1)rk+1
r =0
r1 r =0
r2 r =0
rk+1
1 2 k+1
∞
X
× (−1)m1 r1 α−m1 (s1 +2r1 ) (−1)(m1 +m2 )r2 α−(m1 +m2 )(s2 +2r2 )
m1 ,m2 ,...,mk =1
Pk

· · · (−1) d=1 md rk

Pk
Pk
Pk

× α− d=1 md (sk +2rk )
(−1) d=1 (k+1−d)md rk+1
α− d=1 (k+1−d)md (sk+1 +2rk+1 )

Pk+1 ∞   ∞   ∞  
s
n=1 n
X −s1 r1
X −s2 r2
X −sk+1
= 5 2 (−1) (−1) · · · (−1)rk+1
r =0
r1 r =0
r2 r =0
rk+1
1 2 k+1
∞
X 
m1
× (−1)r1 +···+rk +krk+1 α− s1 +···+sk +ksk+1 +2(r1 +···+rk +krk+1 )

m1 ,m2 ,...,mk =1

 m2
× (−1)r2 +···+rk +(k−1)rk+1 α− s2 +···+sk +(k−1)sk+1 +2(r2 +···+rk +(k−1)rk+1 )

mk
(2.8) × · · · × (−1)rk +rk+1 α− sk +sk+1 +2(rk +rk+1 ) .

Now for any (r1 , . . . , rk ; rk+1 ) ∈ Zk+1

≥0 , we have

∞
m
1

(−1)r1 +···+rk +krk+1 α− s1 +···+sk +ksk+1 +2(r1 +···+rk +krk+1 )
X

m1 ,m2 ,...,mk =1
44 U.K. Dutta and P.K. Ray

r2 +···+rk +(k−1)rk+1 − s2 +···+sk +(k−1)sk+1 +2(r2 +···+rk +(k−1)rk+1 ) m2
× (−1) α



mk
× · · · × (−1)rk +rk+1 α− sk +sk+1 +2(rk +rk+1 )


(−1)r1 +···+rk +krk+1 α− s1 +···+sk +ksk+1+2(r1 +···+rk +krk+1)
=


1− (−1)r1 +···+rk +krk+1 α− s1 +···+sk +ksk+1 +2(r1 +···+rk +krk+1 )


(−1)r2 +···+rk +(k−1)rk+1 α− s2 +···+sk +(k−1)sk+1 +2(r2 +···+rk +(k−1)rk+1 )
×


1 − (−1)r2 +···+rk +(k−1)rk+1 α− s2 +···+sk +(k−1)sk+1 +2(r2 +···+rk +(k−1)rk+1 )


(−1)rk +rk+1 α− sk +sk+1+2(rk +rk+1 )
×··· ×


(1 − (−1)rk +rk+1 α− sk +sk+1 +2(rk +rk+1 )

(−1)r1 +···+rk +krk+1

=
αs1 +···+sk +ksk+1+2(r1 +···+rk +krk+1) − (−1)r1 +···+rk +krk+1

(−1)r2 +···+rk +(k−1)rk+1

×
αs2 +···+sk +(k−1)sk+1 +2(r2 +···+rk +(k−1)rk+1 ) − (−1)r2 +···+rk +(k−1)rk+1

(−1)rk +rk+1
(2.9) × · · · ×
.
αsk +sk+1+2(rk +rk+1 ) − (−1)rk +rk+1

By virtue of (2.8) and (2.9), we have

ζAV F,k (s1 , . . . , sk ; sk+1 )

Pk+1 ∞   ∞  ∞  Pk+1
 
n=1 n
s X −s1 X −s2 X −sk+1
= 5 2 ··· (−1) n=1 rn
r1 r2 rk+1
r1 =0 r2 =0 rk+1 =0

(−1)r1 +···+rk +krk+1

×
αs1 +···+sk +ksk+1+2(r1 +···+rk +krk+1) − (−1)r1 +···+rk +krk+1

(−1)r2 +···+rk +(k−1)rk+1

×
αs2 +···+sk +(k−1)sk+1 +2(r2 +···+rk +(k−1)rk+1 ) − (−1)r2 +···+rk +(k−1)rk+1

(−1)rk +rk+1
(2.10) × · · · ×
.
αsk +sk+1 +2(rk +rk+1) − (−1)rk +rk+1

Thus, this bound is uniform when (s1 , . . . , sk ; sk+1 ) varies over compact subsets
of Ck+1 . The infinite series (2.10) determines the holomorphic function on Ck+1
except for the poles derived from the functions

αsd +···+sk +(k+1−d)sk+1 +2(rd +···+rk +(k+1−d)rk+1 )

(2.11)
−(−1)rd +···+rk +(k+1−d)rk+1 = 0
Analytic properties of the Apostol-Vu multiple Fibonacci ... 45

for d = 1, . . . , k.
Hence, the series (2.10) converges uniformly and absolutely on compact
subsets of Ck+1 not containing any poles of the functions (2.11). Therefore,
ζAV F,k (s1 , . . . , sk ; sk+1 ) can be analytically continued to a meromorphic function
on Ck+1 and its simple poles are on the hyperplanes

sd + · · · + sk + (k + 1 − d)sk+1 = −2(rd + · · · + rk + (k + 1 − d)rk+1 )

πi rd + · · · + rk + (k + 1 − d)rk+1 + 2a
(2.12) + ,
log α
for 1 ≤ d ≤ k, with r1 , . . . , rk+1 ∈ Z≥o and a ∈ Z. This finishes the proof.

3. Residues of Apostol-Vu multiple Fibonacci zeta functions at

poles

We now calculate the residues of ζAV F,k (s1 , . . . , sk ; sk+1 ) at the poles that lie on
the hyperplanes derived from Theorem 3. For 1 ≤ d ≤ k, let us denote

sk+1 (d) = sd + · · · + sk + (k + 1 − d)sk+1 , rk+1 (d) = rd + · · · + rk + (k + 1 − d)rk+1 ,

′
and rk+1 (d) = rd′ + · · · + rk′ + (k + 1 − d)rk+1
′ , with the assumption that
′
sk+1 (d) = 0, rk+1 (d) = 0, rk+1 (d) = 0 for d ≥ k + 1.

We define the residue of the Apostol-Vu multiple Fibonacci zeta functions

ζAV F,k (s1 , . . . , sk ; sk+1 ) along the hyperplanes (2.12) to be the restriction of the
meromorphic function
 

πi
sk+1 (d) + 2rk+1 (d) − rk+1 (d) + 2a ζAV F,k (s1 , . . . , sk ; sk+1 )
log α

to the hyperplanes (2.12).

Theorem 4. Let rd′ , . . . , rk′ , rk+1
′ be non-negative integers for 1 ≤ d ≤ k and
′ ′

πi
ek+1 (d) = −2rk+1 (d) + rk+1 (d) + 2a log α . Then the residue of ζAV F,k (s1 , . . . , sk ;
sk+1 ) at sk+1 (d) = ek+1 (d) is
′
ek+1 (d) (−1)rk+1 (d) Pd−1
   
X −s1 −sk+1
5 2 ··· (−1)( i=1 ri )+(d−k)rk+1
log α r1 rk+1
r1 ,...,rk ,rk+1 ≥0
rk+1 (d)=rk+1 ′ (d)
Pd−1 k
( s )+(d−k)sk+1
i=1 i
Y (−1)rk+1 (l)
×5 2 .
l=1
(αsk+1 (l)+2rk+1 (l) − (−1)rk+1 (l) )
l6=d
46 U.K. Dutta and P.K. Ray

′ ′
Proof. The functions αsk+1 (d)+2rk+1 (d) − (−1)rk+1 (d) are analytic functions which
have simple poles at ek+1 (d) for 1 ≤ d ≤ k.
Now,

(−1)rk+1 (d) (sk+1 (d) − ek+1 (d))

lim
sk+1 (d)→ek+1 (k) αsk+1 (d)+2rk+1 (d) − (−1)rk+1 (d)
1
= (−1)rk+1 (d) Res sk+1 (d)=ek+1 (d)
αsk+1 (d)+2rk+1 (d) − (−1)rk+1 (d)
(−1)rk+1 (d) 1
= r (d)
= .
(−1) k+1 log α log α

Now consider the limit as follows

 
sk+1 (d) X −sd
lim (sk+1 (d) − ek+1 (d))5 2 (−1)rd
sk+1 (d)→ek+1 (d)
rd ,...,rk ,rk+1 =0
rd

(−1)rk+1 (d)
 
−sk+1
··· (−1)rk+1 s (d)+2r (d)
rk+1 (α k+1 k+1 − (−1)rk+1 (d) )
(−1)rk+1 (k)
··· .
(αsk+1 (k)+2rk+1 (k) − (−1)rk+1 (k) )

In the above calculation, after applying the limit, only those terms containing
′
rd , . . . , rk , rk+1 will survive when rk+1 (d) = rk+1 (d) and rest of the terms will
vanish. Thus the above limit becomes
   
ek+1 (d) X −sd rd −sk+1
5 2 (−1) · · · (−1)rk+1
rd rk+1
rd ,...,rk ,rk+1 ≥0
rk+1 (d)=rk+1 ′ (d)

(−1)rk+1 (d+1) (−1)rk+1 (k) 1

s (d+1)+2r (d+1) r (d+1)
· · · s (k)+2r (k) r (k)
.
(α k+1 k+1 − (−1) k+1 ) (α k+1 k+1 − (−1) k+1 ) log α

Therefore, the residue of ζAV F,k (s1 , . . . , sk ; sk+1 ) along sk+1 (d) = ek+1 (d) is

Ressk+1 (d)=ek+1 (d) ζAV F,k (s1 , . . . , sk ; sk+1 )

∞  
s1 +···+sk+1 X −s1
= lim (sk+1 (d) − ek+1 (d))5 2 (−1)r1
sk+1 (d)→ek+1 (d)
r =0
r1
1
∞  
X −sk+1
··· (−1)rk+1
rk+1
rk+1 =0
Analytic properties of the Apostol-Vu multiple Fibonacci ... 47

(−1)rk+1 (1) (−1)rk+1 (k)

× · · ·
(αsk+1 (1)+2rk+1 (1) − (−1)rk+1 (1) ) (αsk+1 (k)+2rk+1 (k) − (−1)rk+1 (k) )
∞     Pk+1
Pk+1
X −s1 −sk+1 (
i=1 i
s )
= ··· lim (sk+1 (d)−ek+1 (d))5 2 (−1)( i=1 ri )
r1 ,...,r =0
r1 rk+1 sk+1 (d)→ek+1 (d)
k+1

k
Y (−1)rk+1 (d)
×
d=1
(αsk+1 (d)+2rk+1 (d) − (−1)rk+1 (d) )
′
(−1)rk+1 (d) Pd−1
   
ek+1 (d) X −s1 −sk+1
=5 2 ··· (−1)( i=1 ri )+(d−k)rk+1
log α r1 rk+1
r1 ,...,rk ,rk+1 ≥0
′
rk+1 (d)=rk+1 (d)
Pd−1 k
( s )+(d−k)sk+1
i=1 i
Y (−1)rk+1 (l)
5 2 .
l=1
(αsk+1 (l)+2rk+1 (l) − (−1)rk+1 (l) )
l6=d

This ends the proof.

References

[1] T.M. Apostol and T.H. Vu, Dirichlet series related to Riemann zeta function, J.
Number Theory 19 (1984) 85–102.
doi:10.1016/0022-314x(84)90094-5
[2] D. Behera, U.K. Dutta and P.K. Ray, On Lucas-balancing zeta function, Acta Com-
ment. Univ. Tartu. Math. 22 (2018) 65–74.
doi:10.12697/ACUTM.2018.22.07
[3] K. Matsumoto, On the analytic continuation of various multiple zeta-functions, in
“Number Theory for the Millennium II, Proc. of the Millennial Conference on Num-
ber Theory” M.A. Bennett et al. (eds.), A K Peters, 2002, pp. 417–440.
[4] K. Matsumoto, On Mordell-Tornheim and other multiple zeta functions, Proceedings
of the Session in Analytic Number Theory and Diophantine Equations, Bonner
Math. Schriften, Vol. 360 (Univ. Bonn, 2003) 17pp.
[5] N.K. Meher and S.S. Rout, Analytic continuation of the multiple Lucas zeta func-
tions, J. Math. Anal. Appl., article in press.
doi:10.1016/j.jmaa.2018.08.063
[6] J. Mehta, B. Saha and G.K. Viswanadham, Analytic properties of multiple zeta
functions and certain weighted variants, an elementary approach, J. Number Theory
168 (2016) 487–508.
doi:10.1016/j.jnt.2016.04.027
[7] L. Navas, Analytic continutation of the Fibonacci Dirichlet series, Fibonacci Quart.
39 (2001) 409–418.
48 U.K. Dutta and P.K. Ray

[8] S.S. Rout and N.K. Meher, Analytic continuation of the multiple Fibonacci zeta
functions, Proc. Japan Acad. Ser. A Math. Sci. 94 (2018) 64–69.
doi:10.3792/pjaa.94.64
[9] S.S. Rout and G.K. Panda, Balancing Dirichlet series and related L-function, Indian
J. Pure Appl. Math. 45 (2014) 943–952.
doi:10.1007/s13226-014-0097-0
[10] D. Zagier, Values of zeta functions and their applications, First European Congress
of Mathematics, Vol. II (Birkhauser, 1994), 210–220.
doi:10.1007/978-3-0348-9112-7
[11] J. Zhao, Analytic continuation of multiple zeta functions, Proc. Amer. Math. Soc.
128 (2000) 1275–1283.
doi:10.2307/119632

Received 25 May 2019

Revised 26 August 2019
Accepted 23 January 2020