Hybrid Priority-Based Latency Optimization for Massive Dependent Tasks

Mobility in edge computing is claimed to be one of the solutions to meet the requirements of massive applications. Nevertheless, the current computational communication schemes are unable to adapt to massive applications. How to define the hybrid priority order for dependent tasks of the applications and optimize the communication and computing latency for massive tasks to enhance the quality of service (QoS) for latency-sensitive network is a great challenge. This article presents a hybrid priority-based latency optimization scheme for massive dependent tasks to address this issue. Firstly, the air-ground integrated mobile edge computing framework is established. The hybrid priority (HP) by taking the six characteristics of the task into account is devised to construct the communication and computing latency model. Then, the Q-learning algorithm is utilized to optimize the latency model. The simulation results showcase that efficiency of the proposed scheme outperforms over other benchmark algorithms in latency optimization.