A lightweight BPSO mechanism for topology reconfiguration in data-driven IIoT plants

Abstract

Wireless Sensor Networks in large-area monitoring applications may be required to deal with emergency events, i.e., events from regions that suddenly require close attention. Nodes in these regions may demand increased data generation rates, and their communication with the base station (BS) should be prioritized. In this sense, a significant challenge is how to reconfigure the network to create paths that offer adequate Quality of Service (QoS) between these sensor nodes and the BSs. Depending on the network’s size, the computational complexity of this problem may become unacceptable; but a meta-heuristic approach could handle it. This work describes a lightweight Binary Particle Swarm Optimization (BPSO) mechanism for the topology reconfiguration of data-driven cluster-tree networks based on IEEE 802.15.4 standard. The goal of this mechanism, called CALiPSO, is to reduce the computational complexity when searching for a suitable network topological configuration that improves network QoS metrics. A simulation assessment and comparison with other approaches were conducted. It was demonstrated that CALiPSO can reconfigure a network of 125 nodes and create an appropriate cluster-tree with fewer cluster-heads in hotspot regions, while improving network QoS. Compared with heuristic and optimization methods, CALiPSO selects around 42% to 55% fewer CHs and forms trees 22% less depth. Concurrently, it diminishes communication delays by about 5% and decreases packet losses by 7.5%, without jeopardizing energy consumption.