On the Identification of Symmetric and Antisymmetric Impulse Responses

Abstract

System identification problems can be efficiently addressed by exploiting some specific characteristics of the impulse responses. In this paper, we focus on the identification of systems that own intrinsic symmetric and antisymmetric properties. Such impulse responses can be modeled based on a combination of bilinear forms, which can be further identified using customized algorithms. In this framework, we propose an iterative Wiener filter tailored for such models. As compared to the conventional Wiener filter, the iterative version is able to achieve significantly better results, especially when a small amount of data is available for the estimation of the required statistics.