An Energy-Efficient Edge Computing Framework for Decentralized Sensing in WSN-Assisted IoT
Vini Gupta, Swades De
Abstract
This paper addresses the problem of decentralized sensor selection in an energy-constrained wireless sensor network-based Internet-of-Things, for monitoring a spatio-temporally varying process. To do so, an adaptive edge computing framework and its variants are proposed which distributedly optimize a critical trade-off between sensing quality and remaining energy of the sensor nodes (SNs). Unlike the existing distributed sensing approaches, the proposed one aims to maintain energy balance among the SNs. The original sensor selection problem is decoupled into multiple sub-problems, each solvable at an edge node elected as head of a coverage region containing a set of SNs. The sub-problem in each coverage region is adapted to variations of the underlying process. In each region, the process is estimated using PCA-SBL (principal component analysis-sparse Bayesian learning) on noisy signal measured by the respective active SNs. Further, to correctly adapt to the process and estimate the signal, a novel logic is designed that indicates requirement of network retraining in the next measurement cycle. The results from extensive simulation studies illustrate improved energy efficiency and network energy balance of the proposed framework over the existing closest competitive centralized and decentralized approaches. The proposed framework is tested on synthetic as well as real data-sets of a sensor network.