Fullerene图的L(p,q)-标号问题
2016-04-11董晓媛马登举
董晓媛,马登举
(1.南通师范高等专科学校数理系,江苏 南通 226000;
2.南通大学理学院,江苏 南通 226007)
Fullerene图的L(p,q)-标号问题
董晓媛1,马登举2
(1.南通师范高等专科学校数理系,江苏 南通 226000;
2.南通大学理学院,江苏 南通 226007)
[摘要]主要研究了一类Fullerene图Fspan的L(2,1)-标号问题及L(1,1)-标号问题,给出了Fspan的L(2,1)-标号数和L(1,1)-标号数的上界分别为7和6.该结果验证了Georges和Mauro猜想与Wegner猜想对于Fullerene图Fspan均成立.
[关键词]L(p,q)-标号,Fullerene图
1预备知识
Fullerene图是一类3-正则3-连通的平面图,且其在平面上的一个嵌入中每个面的边界是5-圈或6-圈.[1-4]
定义图Fm如下:它的顶点集为
边集为
由Fm的定义知,Fm是一类3-正则图,且有m+2个圈,分别为:
(1) 当i=0时,有一个含5个顶点的圈:
C0=u0,1u0,2u0,3u0,4u0,5;
(2) 当i=1,2,…,m时,有一个含10个顶点的圈:
Ci=ui,1ui,2…ui,10ui,1;
(3) 当i=m+1时,有一个含5个顶点的圈:
Cm+1=um+1,1um+1,2um+1,3um+1,4um+1,5.
按照m的奇偶性,Fm的一个平面嵌入如图1与图2所示.易见:Fm是一个3-连通的图,且每个圈的边界是5-圈或6-圈.因此Fm是一类Fullerene图.
图1 当m为奇数时,Fullerene图Fm在平面上的一种嵌入
图2 当m为偶数时,Fullerene图Fm在平面上的一种嵌入
一个图G的L(p,q)-标号,是一个从G的顶点集V(G)到一个非负整数集的一个映射f,使得对G中的任意两个顶点u,v,当d(u,v)=1时,|f(u)-f(v)|≥p;当d(u,v)=2时,|f(u)-f(v)|≥q.这里d(u,v)表示u,v的距离.
一个图G的平方图G2是这样一个图,它的顶点集与G的顶点集相同,两个顶点相邻,当且仅当这两个顶点在G中的距离不大于2.
本文我们主要研究Fullerene图Fm的L(2,1)-标号问题与L(1,1)-标号问题.
2Fullerene图的L(2,1)-标号
定理1图Fm的L(2,1)-标号数λ2,1(Fm)≤7.
证明当m>10时,定义Fullerene图Fm的一个标号f如下:
f(u0,1)=7,f(u0,2)=1,f(u0,3)=6,f(u0,4)=3,f(u0,5)=0;
f(ui,1)=4,f(ui,2)=0,f(ui,3)=3,f(ui,4)=7,f(ui,5)=2,
f(ui,6)=5,f(ui,7)=7,f(ui,8)=4,f(ui,9)=6,f(ui,10)=1,i≡1(mod10);
f(ui,1)=4,f(ui,2)=6,f(ui,3)=1,f(ui,4)=4,f(ui,5)=0,
f(ui,6)=3,f(ui,7)=7,f(ui,8)=2,f(ui,9)=5,f(ui,10)=7,i≡2(mod10);
f(ui,1)=2,f(ui,2)=5,f(ui,3)=7,f(ui,4)=4,f(ui,5)=6,
f(ui,6)=1,f(ui,7)=4,f(ui,8)=0,f(ui,9)=3,f(ui,10)=7,i≡3(mod10);
f(ui,1)=0,f(ui,2)=3,f(ui,3)=7,f(ui,4)=2,f(ui,5)=5,
f(ui,6)=7,f(ui,7)=4,f(ui,8)=6,f(ui,9)=1,f(ui,10)=4,i≡4(mod10);
f(ui,1)=6,f(ui,2)=1,f(ui,3)=4,f(ui,4)=0,f(ui,5)=3,
f(ui,6)=7,f(ui,7)=2,f(ui,8)=5,f(ui,9)=7,f(ui,10)=4,i≡5(mod10);
f(ui,1)=5,f(ui,2)=7,f(ui,3)=4,f(ui,4)=6,f(ui,5)=1,
f(ui,6)=4,f(ui,7)=0,f(ui,8)=3,f(ui,9)=7,f(ui,10)=2,i≡6(mod10);
f(ui,1)=3,f(ui,2)=7,f(ui,3)=2,f(ui,4)=5,f(ui,5)=7,
f(ui,6)=4,f(ui,7)=6,f(ui,8)=1,f(ui,9)=4,f(ui,10)=0,i≡7(mod10);
f(ui,1)=1,f(ui,2)=4,f(ui,3)=0,f(ui,4)=3,f(ui,5)=7,
f(ui,6)=2,f(ui,7)=5,f(ui,8)=7,f(ui,9)=4,f(ui,10)=6,i≡8(mod10);
f(ui,1)=7,f(ui,2)=4,f(ui,3)=6,f(ui,4)=1,f(ui,5)=4,
f(ui,6)=0,f(ui,7)=3,f(ui,8)=7,f(ui,9)=2,f(ui,10)=5,i≡9(mod10);
f(ui,1)=7,f(ui,2)=2,f(ui,3)=5,f(ui,4)=7,f(ui,5)=4,
f(ui,6)=6,f(ui,7)=1,f(ui,8)=4,f(ui,9)=0,f(ui,10)=3,i≡0(mod10).
对于Fullerene图Fm,当m>10时,从F1到F10的L(2,1)-标号以及定义可知:第0圈u0,1,u0,2,u1,3,…,u0,5总是依次标为7,1,6,3,0;第1圈u1,1,u1,2,u1,3,…,u1,9,u1,10依次标为4,0,3,7,2,5,7,4,6,1;第2圈u2,1,u2,2,u2,3,…,u2,9,u2,10依次标为4,6,1,4,0,3,7,2,5,7;照此类推,到第11圈的时候开始与第1圈重复标号,依次为4,0,3,7,2,5,7,4,6,1;第12圈与第2圈重复标号,依次为4,6,1,4,0,3,7,2,5,7;….也就是说,对于Fm,当m>10且m≡imod10,i=0,1,2,…,9时,Fm的第0圈与Fi的第0圈各顶点标号相同,Fm的第m圈与Fi的第i圈各顶点标号相同,Fm的第m+1圈与Fi的第i+1圈各顶点标号相同.而Fm的第1至m-1圈总是标4,0,3,7,2,5,7,4,6,1这10个数(顺序与F10相同).即Fm与Fi的L(2,1)-标号数相同.
这样,上面定义的Fullerene图的标号f是一个7-L(2,1)-标号.因此,Fullerene图的L(2,1)-标号数λ2,1(Fm)≤7.定理证毕.
1992年,Griggs和Yeh得到最大度为Δ的图G的L(2,1)-标号数的下界:
引理1[5]图G包含三个最大度Δ≥2的顶点,且其中一个顶点与另两个顶点相邻,则λ2,1(G)≥Δ+2.
由引理1可知λ2,1(Fm)≥5.
综上,Fm的L(2,1)-标号数5≤λ2,1(Fm)≤7.
2002年,Georges和Mauro[6]提出一个猜想:Petersen图是唯一一个L(2,1)-标号数为9的3-正则图,其他的3-正则图的L(2,1)-标号数总是至多为8.我们发现,对Fullerene图Fm这个猜想成立.
3Fullerene图的L(1,1)-标号
定理2Fm的L(1,1)-标号数λ1,1(Fm)≤6.
证明当m>10时,定义Fullerene图Fm的一个标号f如下:
f(u0,1)=1,f(u0,2)=0,f(u0,3)=3,f(u0,4)=2,f(u0,5)=5;
f(ui,1)=3,f(ui,2)=4,f(ui,3)=2,f(ui,4)=1,f(ui,5)=5,
f(ui,6)=0,f(ui,7)=4,f(ui,8)=3,f(ui,9)=0,f(ui,10)=6,i≡1(mod10);
f(ui,1)=3,f(ui,2)=0,f(ui,3)=6,f(ui,4)=3,f(ui,5)=4,
f(ui,6)=2,f(ui,7)=1,f(ui,8)=5,f(ui,9)=0,f(ui,10)=4,i≡2(mod10);
f(ui,1)=5,f(ui,2)=0,f(ui,3)=4,f(ui,4)=3,f(ui,5)=0,
f(ui,6)=6,f(ui,7)=3,f(ui,8)=4,f(ui,9)=2,f(ui,10)=1,i≡3(mod10);
f(ui,1)=4,f(ui,2)=2,f(ui,3)=1,f(ui,4)=5,f(ui,5)=0,
f(ui,6)=4,f(ui,7)=3,f(ui,8)=0,f(ui,9)=6,f(ui,10)=3,i≡4(mod10);
f(ui,1)=0,f(ui,2)=6,f(ui,3)=3,f(ui,4)=4,f(ui,5)=2,
f(ui,6)=1,f(ui,7)=5,f(ui,8)=0,f(ui,9)=4,f(ui,10)=3,i≡5(mod10);
f(ui,1)=0,f(ui,2)=4,f(ui,3)=3,f(ui,4)=0,f(ui,5)=6,
f(ui,6)=3,f(ui,7)=4,f(ui,8)=2,f(ui,9)=1,f(ui,10)=5,i≡6(mod10);
f(ui,1)=2,f(ui,2)=1,f(ui,3)=5,f(ui,4)=0,f(ui,5)=4,
f(ui,6)=3,f(ui,7)=0,f(ui,8)=6,f(ui,9)=3,f(ui,10)=4,i≡7(mod10);
f(ui,1)=6,f(ui,2)=3,f(ui,3)=4,f(ui,4)=2,f(ui,5)=1,
f(ui,6)=5,f(ui,7)=0,f(ui,8)=4,f(ui,9)=3,f(ui,10)=0,i≡8(mod10);
f(ui,1)=4,f(ui,2)=3,f(ui,3)=0,f(ui,4)=6,f(ui,5)=3,
f(ui,6)=4,f(ui,7)=2,f(ui,8)=1,f(ui,9)=5,f(ui,10)=0,i≡9(mod10);
f(ui,1)=1,f(ui,2)=5,f(ui,3)=0,f(ui,4)=4,f(ui,5)=3,
f(ui,6)=0,f(ui,7)=6,f(ui,8)=3,f(ui,9)=4,f(ui,10)=2,i≡0(mod10).
对于Fullerene图Fm,当m>10时,从F1到F10的L(1,1)-标号以及定义可知:第0圈u0,1,u0,2,u1,3,…,u0,5总是依次标为1,0,3,2,5;第1圈u1,1,u1,2,u1,3,…,u1,9,u1,10依次标为3,4,2,1,5,0,4,3,0,6;第2圈u2,1,u2,2,u2,3,…,u2,9,u2,10依次标为3,0,6,3,4,2,1,5,0,4;依此类推,到第11圈的时候开始与第1圈重复标号,依次为3,4,2,1,5,0,4,3,0,6;第12圈与第2圈重复标号,依次为3,0,6,3,4,2,1,5,0,4;….也就是说,对于Fm,当m>10且m≡i(mod10),i=0,1,2,…,9时,Fm的第0圈与Fi的第0圈各顶点标号相同,Fm的第m圈与Fi的第i圈各顶点标号相同,Fm的第m+1圈与Fi的第i+1圈各顶点标号相同.而Fm的第1至m-1圈总是标3,4,2,1,5,0,4,3,0,6这10个数(顺序与F10相同).即Fm与Fi的L(1,1)-标号数相同.
这样,上面定义的Fullerene图Fm的标号f是一个6-L(1,1)-标号.从而λ1,1(Fm)≤6.定理证毕.
由图的色数及图的L(p,q)-标号的定义,可知图G的平方图的色数χ(G2)与G的L(1,1)-标号数有关,从而我们得出如下结论:
推论1Fm的平方图的色数χ(Fm2)=λ1,1(Fm)+1≤7.
在1977年,Wegner[7]曾猜想一个最大度为3的平面图G的平方图G2的色数χ(G2)≤7.我们的结果表明,对Fullerene图Fm,该猜想成立.
[参考文献]
[1]DOSLIC T. On lower bounds of number of perfect matchings in Fullerene graphs [J]. Mathematical Chemistry,1998,24:359-364.
[2]KARDOS F,KRAL D,MISKUF J. Fullerene graphs have exponentially many perfect matchings [J]. Mathematical Chemistry,2009,46:443-477.
[3]BUHL M,HIRSCH A. Spherical aromaticity of Fullerene [J]. Chem Rev,2001,101:1153-1183.
[4]马海成,汪小玲. 点圈并图的匹配等价圈数[J].东北师大学报(自然科学版),2006,38(4):36-40.
[5]GRIGGS J R,YEH R.K. Labeling graphs with a condition at distance two [J]. SIAM J Discrete Mathematics,1992(5):586-595.
[6]GEORGES J P,MAURO D W. On generalized Petersen graphs labeled with a condition at distance two[J]. SIAM J Discrete Mathematics,2002,259:311-318.
[7]WEGNER G. Graphs with given diameter and a coloring problem[R] Germany:University of Dortmund,1977.
(责任编辑:李亚军)
L(p,q)-labeling of the Fullerene graphs
DONG Xiao-yuan1,MA Deng-ju2
(1.Department of Mathematics and Physics,Nantong Normal College,Nantong 226000,China;2.School of Science,Nantong University,Nantong 226007,China)
Abstract:The L(2,1)-labeling and L(1,1)-labeling of the Fullerene graph Fspanare studied. It is proved that the L(2,1)-labeling number and the L(1,1)-labeling number of Fspanare at most 7 and 6 respectively,which verify the correction of the guesses presented by Georges and Wegner respectively.
Keywords:L(p,q)-labeling;Fullerene graph
[中图分类号]O 157.5[学科代码]110·7470
[文献标志码]A
[作者简介]董晓媛(1984—),女,硕士,讲师,主要从事图的染色问题及其应用研究;通讯作者:马登举(1968—),男,博士,副教授,主要从事图的染色问题及其应用研究.
[基金项目]国家自然科学基金资助项目(11171114).
[收稿日期]2014-06-12
[文章编号]1000-1832(2016)01-0014-04
[DOI]10.16163/j.cnki.22-1123/n.2016.01.004