Energy-optimized and Delay-ensured Task Offloading in High-mobility Vehicle Edge Computing Networks

The rapid growth of the Internet of Vehicles and connected devices has intensified demands for data processing in intelligent transportation systems. While vehicles possess limited onboard computational resources, Vehicular Edge Computing (VEC) networks leverage edge servers to reduce energy consumption and meet latency requirements for delay-sensitive tasks. However, high vehicle mobility poses a significant challenge to effective resource allocation. Most existing studies focus on offloading strategies while overlooking the impact of vehicle mobility. This work formulates a constrained single-objective optimization problem and proposes a hybrid metaheuristic algorithm-Genetic Simulated Annealing Particle Swarm Optimization (GSPSO)-to obtain near-optimal solutions. Experimental results show that GSPSO effectively minimizes energy consumption, achieving reductions of 47.84% and 97.95% compared to Genetic Algorithm and Simulated Annealing Particle Swarm Optimization, respectively.