A Five-dimension MPA Based Fitness Function for Optimizing Energy in IoT Sensor Networks Considering Various Smart City Applications

This paper presents a cutting-edge, Five-dimension MPA Based Fitness Function (FDFF-MPA) designed specifically for Cluster Head (CH) selection in Internet of Things (IoT) networks. With a sharp focus on reducing energy consumption and extending the network’s lifespan, it addresses one of the most pressing challenges in IoT: the limited energy resources of Sensor Nodes (SNs). At the heart of this innovation is a dynamic approach that organizes the network area into several, identically sized hexagonal clusters, each managed by a CH chosen powered by the potent Marine Predator Algorithm (MPA) algorithm. This algorithm selects the most optimal CH based on a novel clustering fitness function that utilizes five vital factors: the remaining energy of SNs, their proximity to the sink, average scale of building occlusions, intra-cluster distance, and $\mathbf{C H}$ rotation times. What sets FDFF-MPA apart is its strategic reduction of setup overhead. Once chosen, a CH stays active for several rounds, operating in a batch-based manner, until its energy dips below a preset threshold. This efficiency continues during data transmission, where a CH consolidates data from SNs and sends it directly to the sink node. Simulation results paint an exciting picture, FDFF-MPA not only significantly prolongs network lifespan but also takes throughput to new heights, outperforming existing protocols in the field.