Variable step-size strategy for distributed parameter estimation of compressible systems in WSNs

Abstract

The Internet of Things (IoT) paradigm depends on sensor networks to collect data and perform information processing as well as communication tasks to attain a common goal. This work presents the formulation and analysis of a sparsity-based variable step-size distributed least-mean-square algorithm based on the diffusion cooperation scheme in wireless sensor networks. In particular, the scheme is applied to compressible systems of which the sparse systems are a special case. The performance of the algorithm is assessed for compressible systems in comparison with existing algorithms in order to showcase the superiority of the proposed algorithm. The proposed scheme is then applied to diffusion-based wireless sensor networks for estimating a compressible system and steady-state analysis is carried out for the proposed scheme.