Sparse Mask Retrieval for Distributed Estimation in Diffusion LMS

Mahdi Shamsi

Session: Frames, Riesz bases, and related topics (Jorge Antezana)

Keywords: Distributed Estimation, DLMS, Sparse Mask, Signal recovery over network, Error Analysis.

TL;DR: This paper analyzes error behavior in Diffusion LMS, considering estimator-level convergence, and proposes an efficient thresholding strategy for support retrieval in masked measurements over a network of estimators.

Abstract: This paper explores a thresholding-based algorithm for Diffusion LMS (DLMS) under limited observability. We analyze estimator convergence in mean and energy, deriving an optimal thresholding strategy. The method effectively handles sparse observations in both time and transform domains. Simulations validate our error analysis and highlight the benefits of a cooperative approach, showing a 10–15 dB improvement in Mean Square Deviation (MSD).

Submission Number: 61