Energy Efficient Multi-Task Offloading in Satellite- Assisted Vehicular Edge Computing Networks: An Improved Soft-Actor-Critic Approach
Lina Wang, Juan Li, Minghui Dai, Haijun Zhang · IEEE Transactions on Vehicular Technology · 2024
With the rapid advancement of internet of vehicles (IoVs), the resource-constrained vehicle terminals cannot satisfy the demands of emerging applications for delay and quality of service (QoS) in remote areas. Low earth orbit (LEO) satellite communication, which is a promising technology for next generation mobile communication, can provide efficient and reliable real-time data transmission and stable service for IoVs. However, the communications between vehicle terminals and satellites cause transmission delay and energy consumption, which have a significant impact on system performance. Mobile edge computing (MEC) can satisfy the latency-sensitive and computation-intensive requirements in vehicular networks. In this paper, we leverage the advantages of LEO satellites to enhance the efficiency of computation offloading in satellite-assisted vehicular edge computing networks. To enhance the effectiveness of computation offloading, we put forward a joint optimization problem to minimize the weighted sum of delay and energy consumption. Despite the mixed-integer nonlinearity of the optimization problem, we formulate the problem as a partially observable markov decision process (POMDP) and present an efficient to attain the optimal solution. Eventually, we conduct simulations to confirm the validity and efficiency of our proposed algorithm. Simulation results indicate that our proposed algorithm can effectively diminish the average delay and energy consumption of the system, compared to other baseline algorithms.