Exploiting observation quality information to enhance the steady-state performance of incremental LMS adaptive networks

In this paper, we investigate the effect of observation quality on the steady-state performance of incremental adaptive networks with LMS learning. We exploit the knowledge of observation quality to adjust the step-size parameter in an adaptive network according to nodes observation quality. We formulate the step-size assignment as a constrained optimization problem and then solve it via Lagrange multipliers approach. We show that applying the optimal step sizes in an incremental adaptive network improves its the steady-state performance. The simulation results are also presented to illustrate the derived theoretical results.