Cooperative optimisation strategy of computation offloading in multi-UAVs-assisted edge computing networks

Abstract

Mobile edge computing has been developed as a promising technology to extend diverse services to the edge of the Internet of Things system. Motivated by the high flexibility and controllability of unmanned aerial vehicles (UAVs), a multi-UAVs-assisted mobile edge computing system is studied to reduce the total consumption of time and energy of terminal equipments. In this system, UAVs act as the computing nodes or relay nodes for process terminal equipment's task. Accordingly, an optimisation problem is formulated to minimise the weighted sum of energy and delay consumption in the edge computing network. To solve the problem, an asynchronous advantage actor–critic (A3C) based deep reinforcement learning algorithm is proposed to obtain the optimal strategy for computation offloading and resource allocation. Experimental results demonstrate that the proposed A3C based algorithm converges fast and outperforms the baseline algorithms in terms of the energy and time consumption of system.