Delay-Aware Communication Service Strategy in USV-assisted Maritime MEC Environment

Unmanned surface vehicles (USVs) play an important role in maritime communication systems, which can improve coverage and communication quality. In this paper, a novel USV-assisted offshore edge computing model is proposed to support time-constrained and low-resource offshore equipment communication and computing. The USV-equipped with computing resources can provide computing services to nearby user equipment. Specifically, we consider the maneuverability of the unmanned vehicle and jointly optimize the user variable, the unmanned vehicle's sailing speed and angle, and the task-offloading ratio to maximize the processing delay under the constraint of energy consumption. Considering the non-convexity, multi-dimensional state space, and continuous action space of the problem, a learning rate decay Proximal Policy Optimization algorithm (DPPO) is proposed. Experimental results show that compared with other algorithms, the proposed algorithm can output the global optimal strategy in processing delay.