Artificial Intelligence-based Optimization of Sink Localization for Self-powered Sensor Networks

Abstract

Wireless sensor networks (WSNs) are mainly communication networks comprised of a large number of miniature sensors using collaboration and self-organization, which have the characteristics of high reliability and low deployment cost. However, the mobile Sink nodes of traditional WSNs have problems such as large network energy consumption and data latency, so this paper introduces the deep learning method, an essential technique of artificial intelligence, and proposes a clustering-based energy optimization CEOMS algorithm by considering the mobility characteristics of Sink nodes and energy consumption-related parameters of sensor nodes and constructing energy consumption functions and performance enhancement functions, respectively; subsequently, we build the standard values of cluster head selection that include energy consumption functions and performance enhancement functions; finally, we calculate the Finally, we calculate the mortality rate of Sink nodes to design the adaptive cluster head self-selection function, and then adaptively adjust the cluster head selection criterion value. The proposed algorithm not only improves the process of cluster head standard value selection and the data transfer efficiency, extends the Sink node network life cycle, reduces the network energy consumption, but also provides a basis for optimizing the localization function of Sink nodes.