Single Bit and Reduced Dimension Diffusion Strategies Over Distributed Networks

TL;DR

This paper proposes innovative diffusion-based adaptive estimation strategies for distributed networks that drastically reduce communication load by transmitting only a single bit or reduced-dimensional data, while maintaining performance.

Contribution

It introduces novel diffusion strategies utilizing random projections for minimal communication, with comprehensive stability analysis and state-space modeling.

Findings

Achieves comparable performance to full information exchange methods

Reduces communication load significantly

Provides stability analysis and state-space description

Abstract

We introduce novel diffusion based adaptive estimation strategies for distributed networks that have significantly less communication load and achieve comparable performance to the full information exchange configurations. After local estimates of the desired data is produced in each node, a single bit of information (or a reduced dimensional data vector) is generated using certain random projections of the local estimates. This newly generated data is diffused and then used in neighboring nodes to recover the original full information. We provide the complete state-space description and the mean stability analysis of our algorithms.