Optimization of AODV Routing in VANETs Using Grasshopper, PSO, and Genetic Algorithms

Abstract

Inter-vehicular networks (VANETs), a subset of mobile ad hoc networks, face significant routing challenges due to rapid topology changes caused by fast-moving nodes. This study proposes a method to enhance the Ad Hoc On-Demand Distance Vector (AODV) routing protocol in VANETs using natureinspired optimization algorithms, namely Grasshopper Optimization Algorithm (GOA), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA). The optimization problem is defined as finding optimal values for 11 AODV control parameters to maximize packet delivery ratio (PDR), minimize average endto-end delay (E2ED), and reduce normalized routing load (NRL). The methodology involves integrating these algorithms with NS-2 simulations, where a VANET scenario with 50 vehicles in a 670×670 m² urban area is modeled. The fitness function, combining PDR, E2ED, and NRL with weights of 0.2, 0.5, and 0.3, respectively, guides the optimization process. Experimental results show that GOA and PSO achieved a PDR of 100% (compared to 97.46% for GA), reduced NRL to 0.34% (from 0.62% for GA), and maintained E2ED at 12.32 ms (compared to 11.05 ms for GA). The fitness function value improved to -0.508 for GOA and PSO, outperforming GA’s -0.514. These findings demonstrate the effectiveness of nature-inspired algorithms in enhancing AODV routing performance in VANETs.