Collaborative Computation Offloading in Multi-UAV-MEC Networks: A Reinforcement Learning Approach

Abstract

To cope with the unprecedented surge in demand for data computing, the promising unmanned aerial vehicle (UAV) assisted mobile edge computing (MEC) has been proposed to enable the network edges to provide closer data processing for users. Hence, data offloading from user to the MEC server will require more efficient. The integration of nonorthogonal multiple access (NOMA) technique with MEC has been shown to provide applications with lower latency and higher energy efficiency. To further enhance offloading performance, in this work, we propose an offloading scheme based on the data division and fusion reinforcement learning (DF-RL) algorithm to handle tasks through multi-user and multi-UAV collaboration. We formulate the optimization problem to minimize the delay and energy consumption of the system, and optimize the offloading strategy through the DF-RL algorithm. Firstly, the data fusion module is used to reduce the processing of repetitive tasks. Secondly, the task is divided into sub-tasks by task segmentation module to better complete the cooperation between UAVs. Finally, reinforcement learning (RL) is used to solve the problem and the optimal offloading strategy decision is obtained. Simulation results show that our algorithm not only has great superiority, but also improves the successful rate of the tasks.