A Note On Eigenvalue Decomposition On Jacket Transform

Abstract

Jacket transforms are defined to be n x n matrices A = (a^) over a field F with the property A4* = nln, where A* is the transpose matrix of elements inverse of Л, i.e., Л* = (<C'), which generalized Hadamard transforms and Center Weighted Hadamard transforms. It has been found that the Jacket transforms are applied to signal and image representation and compression. This paper propose a new eigenvalue decomposition method with Jacket transform. The eigenvalue decomposition methods discussed here may be applied to doubly stochastic processing and the information-theoretic analysis of multiple input multiple output (MIMO) channels.