Density of eigenvalues of random normal matrices with an arbitrary potential, and of generalized normal matrices (Q2473445)

The normal matrix model with an arbitrary potential function is considered, i.e. random normal matrices of some size \(N\) with the spectrum restricted to a compact domain \(D^1\) and probability measure \[ P_N(M)dM=Z_N^{-1}\text{exp}(-N\text{tr}W(M))dM \] where \(dM\) is the measure on the space of normal matrices induced by the Euclidian metric on the space of complex matrices, \(Z_N\) is a normalizing factor and the potential W is a real function on \(\mathbb{C}\) with some regularity condition. Some basic facts about the equilibrium measure are emphasized, such as its existence and uniqueness in the case of a continuous potential on a compact set \(D\subset\mathbb{C}\). The asymptotic eigenvalue distribution in the normal matrix model is provided. The construction of the boundary of the domain is discussed in the case of potentials of the form \(W(z)=\Phi(|z|^2)-P(z)-\overline{P(z)}\), where the function \(\Phi\) is continuous on \([0,\infty)\) and twice continuously differentiable on \([0,\infty)\) and \(P\) is a complex polynomial. The generalized conformal mapping method developed by \textit{V. M. Entov} and \textit{P. I. Etingof} [Eur. J. Appl. Math. 8, No. 1, 23--35 (1997; Zbl 0878.76023)] is used to determine the boundary of the region of eigenvalues explicitly. Some of these results are extended to the case of the generalized normal matrix model. For a fixed collection \(\lambda=(\lambda_1,\dots,\lambda_m)\) of real numbers with \(\sum_{i}\lambda_i=0\) and a domain \(D\), a matrix \(A\) of order \(mN\) over \(\mathbb{C}\) is a generalized normal matrix if \(A=[A_{ij}]\) where \(A_{ij},\;i,j=1,\dots,m\) are \(N\times N\) blocks with \(A_{ij}=0\) unless \(j-i=1\) mod \(m\), the spectrum of \(A_{12}A_{23}\cdots A_{m1}\) is contained in \(D\) and \([A,A^{\dag}]=\text{diag}(\lambda_1I_N,\dots,\lambda_mI_N)\). It is shown that this model is equivalent to the usual normal matrix model in the large \(N\) limit.