Linear combinations of graph eigenvalues

Linear combinations of graph eigenvalues
复制标题

DOI:
10.13001/1081-3810.1242
复制
发表时间:
2006-08
影响因子:
0.7
通讯作者:
V. Nikiforov
V. Nikiforov
中科院分区:
数学4区
文献类型:
--
作者:
V. Nikiforov

文献摘要

被引文献

相似文献

设μ1(G)≥...≥ μn(G)是n阶图G的邻接矩阵的特征值,G是G的补图.设F(G)是由μi(G),μ n-i +1(G),μ iG和μ n-i +1G组成的固定线性组合,1 ≤ i ≤ k。证明了极限limn →∞ 1 nmax {F(G):v(G)= n}总是存在.此外,如果最大值取在某些限制图族上,如“无Kr”图或“r-部”图上,则该陈述仍然成立。29 +<$329 42 n − 25 ≤ max v(G)=n µ1(G)+ µ2(G)≤ 2 <$3 n.这个不等式否定地回答了Gernert的一个问题。
Let µ1 (G) ≥ ...≥ µn (G) be the eigenvalues of the adjacency matrix of a graph G of order n, and G be the complement of G. Suppose F (G) is a fixed linear combination of µi (G), µn−i+1 (G) ,µ i G , and µn−i+1 G , 1 ≤ i ≤ k. It is shown that the limit lim n→∞ 1 n max {F (G ): v (G )= n} always exists. Moreover, the statement remains true if the maximum is taken over some restricted families like "Kr-free" or "r-partite" graphs. It is also shown that 29 + √ 329 42 n − 25 ≤ max v(G)=n µ1 (G )+ µ2 (G) ≤ 2 √ 3 n. This inequality answers in the negative a question of Gernert.