Cooperation gain in incremental LMS adaptive networks with noisy links

Abstract

In this paper, we study the influence of noisy links on the effectiveness of cooperation in incremental LMS adaptive network (ILMS). The analysis reveals the important fact that under noisy communication, cooperation among nodes may not necessarily result in better performance. More precisely, we first define the concept of cooperation gain and compute it for the ILMS algorithm with ideal and noisy links. We show that the ILMS algorithm with ideal links outperforms the non-cooperative scheme for all values of step-size (cooperation gain is always bigger than 1). On the other hand, in the presence of noisy links, cooperation gain is not always bigger than 1 and based on the channel and data statistics, for some values of step-size, non-cooperative scheme outperforms the ILMS algorithm. We presented simulation results to clarify the discussions.