Distributed Low-Rank Adaptive Algorithms Based on Alternating Optimization and Applications

TL;DR

This paper introduces distributed low-rank adaptive algorithms that reduce communication overhead and enhance estimation performance in wireless networks through dimensionality reduction and alternating optimization.

Contribution

It proposes a novel distributed low-rank scheme with joint iterative estimation algorithms based on alternating optimization strategies, improving efficiency and accuracy.

Findings

Significantly reduced communication overhead.

Improved estimation performance in wireless networks.

Validated effectiveness through simulations in sensor networks and smart grids.

Abstract

This paper presents a novel distributed low-rank scheme and adaptive algorithms for distributed estimation over wireless networks. The proposed distributed scheme is based on a transformation that performs dimensionality reduction at each agent of the network followed by transmission of a reduced set of parameters to other agents and reduced-dimension parameter estimation. Distributed low-rank joint iterative estimation algorithms based on alternating optimization strategies are developed, which can achieve significantly reduced communication overhead and improved performance when compared with existing techniques. A computational complexity analysis of the proposed and existing low-rank algorithms is presented along with an analysis of the convergence of the proposed techniques. Simulations illustrate the performance of the proposed strategies in applications of wireless sensor networks and smart grids.