An Optimal Clustering-Based Congestion-Aware Multipath Routing Mechanism in WSN Using Hybrid Optimization and Adaptive Deep Network

Abstract

Wireless Sensor Networks (WSNs) are currently considered an effective distributed sensing technology that boosts the performance of integrated devices and wireless communication. Though WSN offers a novel opportunity for establishing the foundation for utilizing ubiquitous and pervasive computing, it faces some kinds of barriers and difficulties, such as low energy efficiency, data packet loss, and network latency. Especially due to repeatedly altered network design and congestion problems, it influences both network bandwidth utilization as well as efficiency. Therefore, in this work, an effectual congestion-aware multipath routing approach is implemented. The motivation behind this work is to resolve the critical issue of congestion-aware routing in WSNs, which is significant for effective data transmission as well as network performance. The enhancing demand for real-time data processing and transmission in WSNs has resulted in congestion-based issues such as energy depletion, delay, and packet loss. The conventional routing approaches mostly concentrate on optimizing single performance measures, avoiding the complex interplay among factors such as routing congestion, energy consumption, delay, and distance. To resolve these issues, the developed work suggests a Hybrid Heuristic-based Crayfish and Kookaburra Optimization Strategy (HH-CKOS), which comprises the Crayfish Optimization Algorithm (COA) and the Kookaburra Optimization Algorithm (KOA). The developed HH-CKOS algorithm chooses the Cluster Head (CH) from the node's group to enhance the performance of distance, delay, residual energy, energy consumption, load, path loss, and routing congestion. Furthermore, the Adaptive Deep Temporal Convolution Network (ADTCN) model is developed for monitoring the congestion and providing congestion-aware routing, where the parameters are tuned by the developed HH-CKOS algorithm to increase the performance. Finally, the developed system provides a congestion-detected outcome. At last, the performance of the developed system is explored and evaluated with numerous conventional systems and proves its superiority.