Adaptive golf-optimized multi-layer perceptron framework for energy-efficient geographic routing and extended network...

The highly dynamic topology, large scalability, and rapid node mobility in mobile ad-hoc networks (MANETs) pose major challenges to achieving efficient routing and sustaining a long network lifetime. Geographic routing, also referred to as position-based routing, enables data transmission between source and destination nodes based on their spatial coordinates. While such techniques enhance resource utilization and coverage efficiency, they often struggle to deliver optimal routing performance with prolonged network longevity. To overcome these limitations, this study proposes an improved multi-layer perceptron-based golf optimization algorithm for geographic routing (IMLP-GOA-GR) in MANETs. The proposed method exploits each mobile node's locally optimal position in the search space to guide routing decisions, movement, and communication, thereby reducing unnecessary energy expenditure. Simulation results across varying node densities show that IMLP-GOA-GR improves network lifetime by approximately 18-25% and reduces total energy consumption by 20-27% compared to baseline methods. Further, it lowers the average end-to-end delay by 44.44% and improves the packet delivery ratio (PDR) by 37.5%. These outcomes clearly demonstrate the effectiveness of the proposed framework in significantly improving routing efficiency and overall MANET performance.