• 沒有找到結果。

On Symplectic Graphs Modulo

N/A
N/A
Protected

Academic year: 2021

Share "On Symplectic Graphs Modulo"

Copied!
6
0
0

加載中.... (立即查看全文)

全文

(1)

and education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling or

licensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of the

article (e.g. in Word or Tex form) to their personal website or

institutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies are

encouraged to visit:

(2)

Contents lists available atScienceDirect

Discrete Mathematics

journal homepage:www.elsevier.com/locate/disc

On symplectic graphs modulo p

n

Yotsanan Meemark

, Thanakorn Prinyasart

Department of Mathematics, Faculty of Science, Chulalongkorn University, Bangkok, 10330, Thailand

a r t i c l e

i n f o

Article history:

Received 26 September 2010 Received in revised form 20 April 2011 Accepted 5 May 2011

Keywords:

Graph automorphisms Stable range one Symplectic graphs

a b s t r a c t

In this work, we study a family of regular graphs using the 2ν×2νsymplectic group modulo

pn, where p is a prime and n andνare positive integers. We find that this graph is strongly regular only whenν = 1. In addition, we define the symplectic graphs of a symplectic space V over a commutative ring R and show that it is vertex transitive and edge transitive when R has stable range one, which is the case forZpn.

© 2011 Elsevier B.V. All rights reserved.

1. Symplectic graphs modulo pn

Let p be a prime, and let n be a positive integer. For

ν ≥

1, let V(2ν) denote the set of 2

ν

-tuples

(

a1

,

a2

, . . . ,

a

)

of

elements inZpn such that aiis invertible modulo pnfor some i

∈ {

1

,

2

, . . . ,

2

ν}

. Define an equivalence relation

pn on

V(2ν)by

(

a1

,

a2

, . . . ,

a

) ∼

pn

(

b1

,

b2

, . . . ,

b2ν

) ⇔ (

a1

,

a2

, . . . ,

a2ν

) = λ(

b1

,

b2

, . . . ,

b2ν

)

for some

λ ∈

pn. HereZ ×

pn stands for the unit group modulo pn. Write

[

a1

,

a2

, . . . ,

a

]

for the equivalence class of

(

a1

,

a2

, . . . ,

a

)

modulo

pn, and let V(2ν)

pn be the set of all such equivalence classes. Let K

(2ν)be the 2

ν ×

2

ν

nonsingular

alternate matrix overZ/pnZgiven by K(2ν)

=

[

0 Iν

Iν 0

]

2ν×2ν

,

where Iνis the

ν × ν

identity matrix.

The symplectic graph modulo pnon V∼(2pnν)relative to K(

2ν), denoted by Sp(2ν)

pn , is the graph whose vertex set is V(

2ν) ∼pn and with adjacency defined by

[

a1

,

a2

, . . . ,

a

]

is adjacent to

[

b1

,

b2

, . . . ,

b

]

(

a1

,

a2

, . . . ,

a

)

K(2ν)

(

b1

,

b2

, . . . ,

b

)

t

Z × pn

(

a1bν+1

aν+1b1

) + (

a2bν+2

aν+2b2

) + · · · + (

aνb

abν

)

is invertible modulo pn

.

The above adjacency condition is well defined. For, if

(

a1

,

a2

, . . . ,

a

) = λ(

a1

,

a2

, . . . ,

a2ν

)

and

(

b1

,

b2

, . . . ,

b

) =

µ(

b′ 1

,

b ′ 2

, . . . ,

b ′ 2ν

)

for some

λ, µ ∈

Z × pn, then

λµ ∈

Z × pnand

(

a1bν+1

aν+1b1

) + (

a2bν+2

aν+2b2

) + · · · + (

aνb

abν

)

=

λµ[(

a1bν+1

aν+1b1

) + (

a2bν+2

aν+2b2

) + · · · + (

aνb2ν

a2νbν

)].

Corresponding author.

E-mail addresses:yotsanan.m@chula.ac.th(Y. Meemark),thanakorn_dpst@hotmail.com(T. Prinyasart). 0012-365X/$ – see front matter©2011 Elsevier B.V. All rights reserved.

(3)

A simple counting shows that

|

V(2ν)

| =

(

pn

)

(

pn−1

)

and

|

V(2ν)

pn

| =

(pn)2ν−(pn−1)2ν

pnpn−1 . Furthermore, we have the following results.

Theorem 1.1. Let p be a prime, and let n and

ν

be positive integers. The symplectic graph Sp(p2nν)is

(

pn

)

2ν−1-regular and every

two adjacent vertices of Sp(p2nν)has

(

pn

)

2ν−2

(

pn

pn−1

)

common neighbors.

Proof. Assume that

[

a1

,

a2

, . . . ,

a

]

and

[

b1

,

b2

, . . . ,

b

]

are adjacent. Since

(

a1

,

a2

, . . . ,

a

) ∈

V(2ν), there exists an i

∈ {

1

,

2

, . . . ,

2

ν}

such that aiis invertible modulo pn. If i

ν

, then

bν+i

=

a

−1

i

(

r

+

(

aν+1b1

a1bν+1

) + (

aν+2b2

a2bν+2

) + · · · + (

aν+i−1bi−1

ai−1bν+i−1

)

+

aν+ibi

+

(

aν+i+1bi+1

ai+1bν+i+1

) + · · · + (

abν

aνb

))

for some r

pnand if i

ν +

1, then

bi−ν

=

a −1

i

((

a1bν+1

aν+1b1

) + (

a2bν+2

aν+2b2

) + · · · + (

ai−1−νbi−1

ai−1bi−1−ν

)

+

ai−νbi

+

(

ai+1−νbi+1

ai+1bi+1−ν

) + · · · + (

aνb

abν

) −

s

)

for some s

pn. Therefore, there are(

pn)2ν−1(pnpn−1) pnpn−1

=

(

p

n

)

2ν−1classes adjacent to the vertex

[

a

1

,

a2

, . . . ,

a

]

, and hence

Sp(p2nν)is

(

pn

)

2ν−1-regular.

Next, let

[

x1

,

x2

, . . . ,

x

]

be a common neighbor of

[

a1

,

a2

, . . . ,

a

]

and

[

b1

,

b2

, . . . ,

b

]

. Then

(

a1xν+1

aν+1x1

) + (

a2xν+2

aν+2x2

) + · · · + (

aνx

axν

) =

r ′ (1.1) and

(

b1xν+1

bν+1x1

) + (

b2xν+2

bν+2x2

) + · · · + (

bνx

bxν

) =

s′ (1.2) for some r

,

s

Z×

pn. Since p-aiand we may assume without loss of generality that i

ν

, from Eq.(1.1)we have

xν+i

=

a −1 i

(

r

+

(

aν+1x1

a1xν+1

) + (

aν+2x2

a2xν+2

) + · · · + (

aν+i−1xi−1

ai−1xν+i−1

)

+

aν+ixi

+

(

aν+i+1xi+1

ai+1xν+i+1

) + · · · + (

axν

aνx

)).

Substituting the value of xν+iinto ai

×

(1.2)and subtracting from bi

×

(1.1)give

ν

j=1

(

aibν+j

aν+ibi

)

xj

+

ν

j=1

(

aibj

ajbi

)

xν+j

=

ais

bir

.

Suppose that p

|

(

aibj

ajbi

)

for all j

∈ {

1

,

2

, . . . ,

2

ν}

. Then we can prove that p

|

(

akbl

albk

)

for all k

,

l

∈ {

1

,

2

, . . . ,

2

ν}

. This implies p

| [

(

a1b2

a2b1

) + (

a3b4

a4b3

) + · · · + (

a2ν−1b

ab2ν−1

)]

which contradicts

[

a1

,

a2

, . . . ,

a

]

is adjacent

to

[

b1

,

b2

, . . . ,

b

]

. Thus, there exists a j

∈ {

1

,

2

, . . . ,

2

ν}

such that

(

aibj

ajbi

) ∈

Z

×

pn. Hence, there are

(

pn

)

2ν−2

(

pn

pn−1

)(

pn

pn−1

)

pn

pn−1

=

(

p

n

)

2ν−2

(

pn

pn−1

)

classes of common neighbors of adjacent vertices

[

a1

,

a2

, . . . ,

a

]

and

[

b1

,

b2

, . . . ,

b

]

. 

Corollary 1.2. The symplectic graph Sp(p2)is a complete graph.

A strongly regular graph with parameters

(v,

k

, λ, µ)

is a k-regular graph on

v

vertices such that for every pair of adjacent vertices there are

λ

vertices adjacent to both, and for every pair of non-adjacent vertices there are

µ

vertices adjacent to both. Therefore,Theorem 1.1shows that our symplectic graphs satisfy three parameters

v =

(pn)2ν−(pn−1)2ν

pnpn−1

,

k

=

(

p n

)

2ν−1

and

λ = (

pn

)

2ν−2

(

pn

pn−1

)

. Unfortunately, the last parameter

µ

is fulfilled only when

ν =

1. Recall from Chapter 10 of [1] that the eigenvalues of a non-complete connected strongly regular graph with parameters

(v,

k

, λ, µ)

are (λ−µ)±

√ ∆

2 and k

and the corresponding multiplicities are12

v −

1

±

(v−1)(µ−λ)−√ 2k

and 1, where

∆ =

(λ − µ)

2

+

4

(

k

µ) >

0.

Theorem 1.3. For n

2, the symplectic graph Sp(p2n) is strongly regular with parameters

(

pn

+

pn−1

,

pn

,

pn

pn−1

,

pn

)

.

Consequently, its eigenvalues are

pn−1

,

0 and pnwith multiplicities pn

+

pn−1

p

1

,

p and 1, respectively.

To prove this theorem, it remains to verify the final parameter

µ

. This will be a consequence of the following two lemmas. The first one talks about the representatives for the vertex set V∼(2pn), which can be partitioned into p

+

1 sets of the same

(4)

Lemma 1.4. Let p be a prime and n a positive integer. For k

∈ {

0

,

1

, . . . ,

p

1

}

, let Ak

= {[

1

,

pi

+

k

] :

i

∈ {

0

,

1

,

2

, . . . ,

pn

−1

1

}}

,

and let A

p

= {[

j

,

1

] :

0

j

<

pnand p

|

j

}

.

Then

|

Ak

| =

pn−1for all k

∈ {

0

,

1

, . . . ,

p

}

and the collection

{

Ak

:

k

∈ {

0

,

1

, . . . ,

p

}}

is a partition of V∼(2pn), that is, Ak

Al

= ∅

for all k

,

l

∈ {

0

,

1

, . . . ,

p

}

and k

̸=

l, and V(2)

pn

=

p k=0Ak.

The second lemma gives a necessary and sufficient condition for adjacent vertices of Sp(p2n), and a conclusion for

Theorem 1.3. Observe that for k

∈ {

0

,

1

, . . . ,

p

1

}

, [

1

,

pi

+

k

]

is not adjacent to

[

1

,

pj

+

k

]

for all i

,

j

∈ {

0

,

1

, . . . ,

pn−2

}

and

[

i

,

1

]

is not adjacent to

[

j

,

1

]

for all 0

i

,

j

<

pn

,

p

|

i and p

|

j. Moreover, we prove the following lemma.

Lemma 1.5. Let p be a prime and n a positive integer. Let k

,

l

∈ {

0

,

1

, . . . ,

p

}

, [

a1

,

a2

] ∈

Akand

[

b1

,

b2

] ∈

Al. Then k

̸=

l if and

only if

[

a1

,

a2

]

is adjacent to

[

b1

,

b2

]

. In addition, every two non-adjacent vertices of Sp( 2)

pn have pncommon neighbors.

Proof. The above observation shows the necessity. To prove the sufficiency, we may assume without loss of generality that

k

<

l. If l

=

p, then

[

a1

,

a2

] = [

1

,

pi

+

k

]

for some i

∈ {

0

,

1

,

2

, . . . ,

pn−1

1

}

and

[

b1

,

b2

] = [

j

,

1

]

for some 0

j

<

pn

and p

|

j, so

[

a1

,

a2

]

is not adjacent to

[

b1

,

b2

]

caused by p -

(

1

(

pi

+

k

)

j

)

. Next, we suppose that k

<

l

<

p. Then

[

a1

,

a2

] = [

1

,

pi

+

k

]

and

[

b1

,

b2

] = [

1

,

pj

+

l

]

for some i

,

j

∈ {

0

,

1

,

2

, . . . ,

pn−1

1

}

. Since 0

k

<

l

<

p

,

p does not divide p

(

j

i

) +

l

k, and hence

[

a1

,

a2

]

is not adjacent to

[

b1

,

b2

]

as desired.

Finally, we show the final statement which completes the proof ofTheorem 1.3. Suppose that

[

a1

,

a2

]

and

[

b1

,

b2

]

are

non-adjacent vertices. Then both of them belong to Alfor some l

∈ {

0

,

1

, . . . ,

p

}

. Thus, the set of their common neighbors is V∼(2pn) rAlwhich consists of pnelements. 

Lemma 1.5also gives an immediate corollary.

Corollary 1.6. Let p be a prime and n a positive integer. The symplectic graph Sp(p2n)is a

(

p

+

1

)

-partite graph with the partition

{

Ak

:

k

∈ {

0

,

1

, . . . ,

p

}}

. As a result, its chromatic number is p

+

1.

Since Sp(p2n) is a complete

(

p

+

1

)

-partite regular graph whose vertex set is partitioned by the collection

{

Ak

:

k

{

0

,

1

, . . . ,

p

}}

, an automorphism of this graph corresponds with a permutation of

{

A0

,

A1

, . . . ,

Ap

}

, and a permutation of vertices in Akfor all k

=

0

,

1

, . . . ,

p. Hence, we have the following theorem.

Theorem 1.7. Let p be a prime and n a positive integer. Then

|

Aut Sp(p2n)

| =

(

pn

−1

!

)

p+1

(

p

+

1

)!

.

We conclude this section with an example showing that if

ν ≥

2, then the parameter

µ

for the symplectic graph Sp(p2nν)

is not fulfilled, and thus it is not strongly regular.

Example 1.8. Let

ν ≥

2. Consider three vertices

a

= [

1

,

0

,

0

, . . . ,

0

]

,b

= [

1

,

0

, . . . ,

0

,

p

,

0

, . . . ,

0

]

, where p is at the

ν +

1st entry, and

c

= [

1

,

1

,

0

,

0

, . . . ,

0

]

in V∼(2pnν). It is easy to see that

a is not adjacent to

b and

b is not adjacent to

c. Assume that

[

x1

,

x2

, . . . ,

x

]

is a common neighbor for

a and

b. Then p must divide xν+1and other xi’s are free. This implies that there are

(

pn

pn−1

)

pn(2ν−1)distinct classes of common neighbors for

a and

b. On the other hand, suppose that

[

y

1

,

y2

, . . . ,

y

]

is a

common neighbor for

b and

c. Thus we must have gcd

(

yν+1

,

p

) =

gcd

(

yν+1

+

yν+2

,

p

) =

1 and other yj’s are free. This yields

(

pn

pn−1

)

2pn(2ν−2)distinct classes of common neighbors forb and

c. Hence, there are two pairs of non-adjacent vertices

such that the numbers of common neighbors of them are not the same, and so the symplectic graph Sp(p2nν)is not strongly

regular.

Remark. Tang and Wan [3] worked on the general symplectic graph Sp

(

2

ν,

q

)

over the finite fieldFq. This graph is strongly regular with parameters

q2qν11

,

q2ν−1

,

q2ν−2

(

q

1

),

q2ν−2

(

q

1

)

. Their proof used orthogonal complements and matrix

theory over finite fields. In contrast, our symplectic graph is defined analogously over the commutative ringZpn and our

proof uses combinatorial method.

2. Vertex transitive and edge transitive

Let R be a commutative ring and let V be a free R-module of R-dimension n, where n

2. Assume that we have a function

β :

V

×

V

R which is R-bilinear,

β(⃗

x

, ⃗

x

) =

0 for all

x

V and the R-module morphism from V to V

=

HomR

(

V

,

R

)

given by

x

→

β(·, ⃗

x

)

is an isomorphism. We call the pair

(

V

, β)

a symplectic space.

A vector

x in V is said to be unimodular if there is an f in Vwith f

(⃗

x

) =

1; equivalently, if

x

=

α

1

b1

+ · · · +

α

nb

nwhere

{⃗

b1

, . . . , ⃗

bn

}

is a basis for V and

α

1

, . . . , α

n

R, then the ideal

1

, . . . , α

n

) =

R. If

x is unimodular, then the line Rx is a free

(5)

Lemma 2.1. Let

x and

y be unimodular vectors in V . Then R

x

=

Ry if and only if

x

=

λ⃗

y for some

λ ∈

R×.

Proof. Assume that R

x

=

R

y. Then

x

=

λ⃗

y for some

λ ∈

R. Since

x is unimodular, there is an f

Vsuch that f

(⃗

x

) =

1. Thus, 1

=

f

(⃗

x

) =

f

(λ⃗

y

) = λ

f

(⃗

y

)

, so

λ

is a unit in R. The converse is clear. 

A hyperbolic pair

{⃗

x

, ⃗

y

}

is a pair of unimodular vectors in V with the property that

β(⃗

x

, ⃗

y

) =

1. The module R

x

R

y is called a hyperbolic plane. Let

(

V

, β)

be a symplectic space. An R-module automorphism T on V is an isometry on V if

β(

T

(⃗

x

),

T

(⃗

y

)) = β(⃗

x

, ⃗

y

)

for all

x

, ⃗

y

V . The group of isometries on V is called the symplectic group of V over R and denoted by SpR

(

V

)

.

Define the graphGSpR(V)with the vertex set is the set of lines

{

R

x

: ⃗

x is a unimodular in V

}

and with adjacency given by

R

x is adjacent to R

y if and only if

β(⃗

x

, ⃗

y

) ∈

R×(equivalently,

β(⃗

x

, ⃗

y

) =

1). We callGSpR(V), the symplectic graph of V over R. To see that this adjacency condition is well defined, let

x1

, ⃗

x2

, ⃗

y1and

y2be unimodular vectors in V . Assume that R

x1

=

R

x2

and Ry

1

=

R

y2. ByLemma 2.1,

x1

=

λ⃗

x2and

y1

=

λ

y

2for some

λ, λ

R×, and hence we have

β(⃗

x1

, ⃗

y1

) ∈

R ×

β(λ⃗

x2

, λ

y2

) ∈

R ×

λλ

β(⃗

x2

, ⃗

y2

) ∈

R ×

β(⃗

x2

, ⃗

y2

) ∈

R ×

.

A commutative ring R is said to have stable range one if for all

α, β ∈

R with

(α, β) =

R, there exists a

δ

in R such that

α + δβ ∈

R×. Kirkwood and McDonald (Theorem 2.8 of [2]) showed the following results.

Lemma 2.2 ([2]). Let R be a commutative ring with stable range one and 2 a unit. Let V be a symplectic space over R. Then,

SpR

(

V

)

acts transitively on unimodular vectors and on hyperbolic planes.

We shall apply the above lemma to verify that our symplectic graph is vertex transitive and edge transitive (that is, its automorphism group acts transitively on the vertex set and the edge set, respectively) when R has stable range one and 2 is a unit.

Theorem 2.3. Let R be a commutative ring with stable range one and 2 a unit. Then the symplectic graphGSpR(V) is vertex

transitive and edge transitive.

Proof. Observe that for any automorphism T of V , we have the induced automorphism

σ

Ton the vertex set of the symplectic

graphGSp(V)given by

σ

T

:

R

x

→

RT

(⃗

x

)

for all unimodular vectors

x

V . Let

x andy be unimodular vectors in V . By

Lemma 2.2, there is an automorphism T

SpR

(

V

)

such that T

(⃗

x

) = ⃗

y. Thus, we have

σ

T

AutGSpR(V)and

σ

T

:

R

x

→

RT

(⃗

x

) =

R

y.

Next, let

x1

, ⃗

x2

, ⃗

y1

, ⃗

y2be unimodular vectors in V such that Rx

1is adjacent to R

y1and R

x2is adjacent to R

y2. We may

assume that

β(⃗

x1

, ⃗

y1

) =

1

=

β(⃗

x2

, ⃗

y2

)

. That is,

{⃗

x1

, ⃗

y1

}

and

{⃗

x2

, ⃗

y2

}

are hyperbolic pairs. Again, byLemma 2.2, there exists

an automorphism T

SpR

(

V

)

such that T

(⃗

x1

) = ⃗

x2and T

(⃗

y1

) = ⃗

y2. Therefore,

σ

T

AutGSpR(V)sends Rx

1to R

x2and R

y1to R

y2as desired. 

Lemma 2.4. Let p be a prime, and let n be a positive integer. Then,Zpnhas stable range one.

Proof. Let

α, β ∈

Zpnbe such that

(α, β) =

Zpn. If p

|

α

, then p-

β

, so p-

(α +

1

β)

and if p-

α

, then p-

(α +

0

β)

. Hence,

Zpnhas stable range one. 

Back to the symplectic graph studied in Section1. If p is a prime,

ν

and n are positive integers, R

=

Zpn, V

=

Z2pνn and

β : (⃗

x

, ⃗

y

) → ⃗

xK(2ν)

yt for all

x

, ⃗

y

V , then it follows thatGSpR(V)is isomorphic to Sp(

2ν)

pn . Thus,Theorem 2.3andLemma 2.4

show that Sp(p2nν)is vertex transitive and edge transitive when p is odd. We record this result in the following theorem. Theorem 2.5. Let p be an odd prime, and let n and

ν

be positive integers. The symplectic graph Sp(p2nν)is vertex transitive and

edge transitive.

Recall from Proposition 2.10 of [2] that the center of SpR

(Z

2pnν

)

is

I

}

, and so we have the quotient group PSpR

(Z

2pνn

) =

SpR

(Z

pn

)/{±

I

}

, which is called the projective symplectic group of degree 2

ν

overZpn. Our final theorem tells us that we may

regard PSpR

(Z

2pνn

)

as a subgroup of the automorphism group of the symplectic graph Sp(

2ν)

pn .

Theorem 2.6. For any T1and T2in SpR

(Z

pn

), σ

T1

=

σ

T2 if and only if T1

= ±

T2, where the induced automorphism

σ

T

,

T

SpR

(Z

2pνn

)

, on the vertex set of the symplectic graph Sp(

2ν)

pn is given by

σ

T

:

Zpn

x

→

ZpnT

(⃗

x

)

(6)

Proof. It is clear that

σ

T1

=

σ

T2if T1

= ±

T2. For the opposite direction, suppose that

σ

T1

=

σ

T2. That is,ZpnT1

(⃗

x

) =

ZpnT2

(⃗

x

)

for all unimodular vectors

x. Thus, byLemma 2.1, for any unimodular vector

x

Z2pνn

,

T1

(⃗

x

) = λ

T2

(⃗

x

)

for some

λ ∈

Z

×

pn.

In particular, if

{⃗

e1

, ⃗

e2

, . . . , ⃗

e

}

is the standard basis forZ2pnν, we have for i

∈ {

1

,

2

, . . . ,

2

ν},

T1

(⃗

ei

) = λ

iT2

(⃗

ei

)

for some

λ

i

Z

×

pn. Let i

∈ {

1

,

2

, . . . , ν}

. Since T1and T2are isometries onZ2pνn,

eiK(2ν)

eν+t i

=

T1

(⃗

ei

)

K(2ν)T1

(⃗

eν+i

)

t

=

λ

i

λ

ν+iT2

(⃗

ei

)

K(2ν)T2

(⃗

eν+i

)

t

=

λ

i

λ

ν+i

eiK(2ν)

etν+i which implies

λ

i

λ

ν+i

=

1. In addition,

eiK(2ν)

1t

=

T1

(⃗

ei

)

K(2ν)T1

(⃗

1

)

t

=

λ

i

λ

T2

(⃗

ei

)

K(2ν)T2

(⃗

1

)

t

=

λ

i

λ⃗

eiK(2ν)

1t

,

and

(−⃗

eν+i

)

K(2ν)

1t

=

T1

(−⃗

eν+i

)

K(2ν)T1

(⃗

1

)

t

=

λ

ν+i

λ

T2

(−⃗

eν+i

)

K(2ν)T2

(⃗

1

)

t

=

λ

ν+i

λ(−⃗

eν+i

)

K(2ν)

1t

,

where

1

= [

1

,

1

, . . . ,

1

]

and T1

(⃗

1

) = λ

T2

(⃗

1

)

for some

λ ∈

Z

×

pn. Thus,

λ

1

=

λ

2

= · · · =

λ

=

λ

−1

.

Recall that

λ

i

λ

ν+i

=

1 for all i

∈ {

1

,

2

, . . . , ν}

, we finally reach T1

= ±

T2as desired. 

Acknowledgments

This work was finished while the first author was visiting the National Center for Theoretical Sciences (NCTS), Mathematics Division, Hsinchu, Taiwan under his exchanged program sponsored by Chulalongkorn University. He expresses his gratitude to NCTS hospitality and academic funding of Chulalongkorn University. The second author would like to thank the Development and Promotion of Science and Technology Talents Project for partially support. Last but not least, we wish to thank the anonymous referees for valuable comments which corrected and improved the quality of the paper.

References

[1] C. Godsil, G. Royle, Algebraic Graph Theory, Springer, New York, 2001.

[2] B. Kirkwood, B.R. McDonald, The symplectic group over a ring with one in its stable range, Pacific J. Math. 92 (1) (1981) 111–125. [3] Z. Tang, Z. Wan, Symplectic graphs and their automorphisms, European J. Combin. 27 (2006) 38–50.

參考文獻

相關文件

• Content demands – Awareness that in different countries the weather is different and we need to wear different clothes / also culture. impacts on the clothing

Assume that the partial multiplicities of purely imaginary and unimodular eigenvalues (if any) of the associated Hamiltonian and symplectic pencil, respectively, are all even and

Courtesy: Ned Wright’s Cosmology Page Burles, Nolette &amp; Turner, 1999?. Total Mass Density

Monopolies in synchronous distributed systems (Peleg 1998; Peleg

From all the above, φ is zero only on the nonnegative sides of the a, b-axes. Hence, φ is an NCP function.. Graph of g functions given in Example 3.9.. Graphs of generated NCP

Corollary 13.3. For, if C is simple and lies in D, the function f is analytic at each point interior to and on C; so we apply the Cauchy-Goursat theorem directly. On the other hand,

Corollary 13.3. For, if C is simple and lies in D, the function f is analytic at each point interior to and on C; so we apply the Cauchy-Goursat theorem directly. On the other hand,

The vertex-cover problem is to find a vertex cover of minimum size in a given undirected graph. • 此問題的decision版本為NP-Complete