Energy-Efficient Clustering Algorithm-Based Routing Protocol by Optimized Machine Learning Algorithms in WSN

Pocket-friendly, small sensor nodes make up a wireless sensor network (WSN). The sensor nodes are designed to collect and transmit data from their environment to the base station (BS). The lifespan of the network is impacted by sensor nodes, which also consume more energy when sending data. Additional energy-related limitations on WSNs include restricted computation, high setup complexity, storage, clustering, and communication capability. The primary characteristics of WSNs are energy efficiency and lifetime extension, which are handled by clustering and routing strategies. Hence, an energy-efficient clustering-based routing (EECR) protocol for WSN is proposed in this study. Initially, the sensor nodes are clustered using the Fuzzy K-Medoids method. Then, the hybrid approach of mayfly and moth flame optimization (HMFMFO) is introduced for optimal cluster head (CH) selection. The hybrid algorithm improves the global search behavior of moth flames by MF and achieves optimal positioning of the CH. Finally, the optimized artificial neural network (Opt-ANN) is used for optimal route selection. The particle swarm optimization (PSO) is used to optimize ANN in order to find the quickest path while dynamically decreasing network overhead. The packet delivery ratio (PDR), efficiency, packet delay, power consumption and network lifespan are used to assess the effectiveness of the suggested strategy. The proposed method achieves with 22.49[Formula: see text]ms delay and energy consumption of 0.0416[Formula: see text]mJ, which is superior to the existing methods.