Limit distribution of eigenvalues for random Hankel and Toeplitz band matrices

From MaRDI portal
(Redirected from Publication:662873)



Abstract: Consider real symmetric, complex Hermitian Toeplitz and real symmetric Hankel band matrix models, where the bandwidth bNaiy but bN/Nob, bin[0,1] as Noinfty. We prove that the distributions of eigenvalues converge weakly to universal, symmetric distributions gammaT(b) and gammaH(b). In the case b>0 or b=0 but with the addition of bNgeqCN1/2+epsilon0 for some positive constants epsilon0 and C, we prove almost sure convergence. The even moments of these distributions are the sum of some integrals related to certain pair partitions. In particular, when the bandwidth grows slowly, i.e. b=0, gammaT(0) is the standard Gaussian distribution and gammaH(0) is the distribution |x|exp(−x2). In addition, from the fourth moments we know that the gammaT(b)'s are different for different b's, the gammaH(b)'s different for different bin[0,1/2] and the gammaH(b)'s different for different bin[1/2,1].


The authors consider real symmetric and complex Hermitian Toeplitz, and real symmetric Hankel band matrix models under some bandwidth conditions. It is proved that the distributions of eigenvalues converge weakly to universal symmetric distributions \(\gamma T(b)\) and \(\gamma H(b)\). In some cases the almost sure convergence is proved. The even moments of these distributions are sums of some integrals. If \(b=0\), the corresponding distribution is the standard Gauss distribution. It is proved also that for different \(b\) the distributions \(\gamma T(b)\) are different. A similar result is true for the distribution \(\gamma H(b)\) but only for \(b\) in the interval \([0,1]\).




Cited in
(21)








This page was built for publication: Limit distribution of eigenvalues for random Hankel and Toeplitz band matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q662873)