Universality and the circular law for sparse random matrices (Q433915)

It is assumed that \(A_n\) is an \(n\) by \(n\) complex matrix with i. i. d. random entries, each having zero mean and unit variance. The character of the distribution of \((1/\sqrt n)A_n\) which approaches the uniform distribution on the unit disk as \(n\) goes to infinity is described by the circular law. In this paper the universality of a distribution is proved for the case of the sparse random \(n\) by \(n\) matrices whose entries are nonzero with probability \(1/n^{1- \alpha}\), where \(0 < \alpha < 1\) is any constant. The universality principle asserts that the distribution of the eigenvalues of a random matrix with i. i. d. zero mean and unit variance entries does not depend on the underlying structure of the random entries. As the consequence of the sparse universality principle it follows that the circular law is valid for sparse matrices so long as the entries have zero mean and unit variance.