SAGVN Task Offloading and Resource Allocation with Spectrum Resource Constraints

Space-air-ground integrated vehicular network (SAGVN) expands the communication range and satisfy people’s needs for long-distance communication. However, it is difficult to realize low-delay and low-energy task processing by solely relying on the computational resources of terrestrial networks, and limited spectrum resources further limits SAGVN development. To address these problems, an SAGVN task offloading and resource allocation (STOR) algorithm is proposed. First, an SAGVN computing offloading system is designed, which enables vehicle user terminals to offload tasks to the peripheral vehicles or mobile edge computing, unmanned aerial vehicle, and low-orbit satellite server for processing and efficiently use spectrum resources through channel multiplexing. Second, an optimization problem with the objective of minimizing the average delay and energy consumed of all vehicles is established, and the optimization problem is transformed into offloading and computing resource allocation subproblems. Finally, the tanh-genetic and Sine mapping particle swarm optimization (PSO) algorithms are employed to solve subproblems, which satisfy the requirements of vehicles. Simulation results show that the STOR algorithm can reduce the total task processing delay by 7.13%, 3.04%, and 21.08%, and the total task energy consumed by 3.87%, 55.21%, and 15.10%, respectively, compared with the random channel allocation, average resource allocation, and simulated annealing-PSO algorithms.