PERFORMANCE ASSESSMENT OF THE SAGE ALGORITHM IN UNKNOWN NOISE FIELDS

Kurzfassung

In this paper, we explore the performance of the Space Alternating Generalized Expectation-Maximization (SAGE) algorithm in estimating the parameters of multipath components in wireless propagation channels. The main goal of this contribution is to extend the theoretical framework of the algorithm to account for the general case of an arbitrary unknown noise field with non-uniform structure. This scenario is of high practical importance since it encounters in several modern applications and, more importantly, inherently encompasses the common case of white uniform additive noise. In the sequel, it is thoroughly demonstrated how the accuracy and convergence rate of the estimators are affected by the spatio-temporal parameters of the incoming wavefronts.