Multiple Sources Localization with Sparse Recovery under Log-normal Shadow Fading

TL;DR

This paper proposes a novel RSS-based multi-source localization method in wireless sensor networks that accounts for shadow fading effects, using sparse recovery and maximum likelihood estimation to improve accuracy.

Contribution

It introduces a sparse recovery and weighted averaging approach combined with Fenton-Wilkinson approximation for better localization under shadow fading.

Findings

Higher localization accuracy demonstrated in simulations

Outperforms existing methods in shadow fading scenarios

Effective handling of unknown transmitted power

Abstract

Localization based on received signal strength (RSS) has drawn great interest in the wireless sensor network (WSN). In this paper, we investigate the RSS-based multi-sources localization problem with unknown transmitted power under shadow fading. The log-normal shadowing effect is approximated through Fenton-Wilkinson (F-W) method and maximum likelihood estimation is adopted to optimize the RSS-based multiple sources localization problem. Moreover, we exploit a sparse recovery and weighted average of candidates (SR-WAC) based method to set up an initiation, which can efficiently approach a superior local optimal solution. It is shown from the simulation results that the proposed method has a much higher localization accuracy and outperforms the other