Energy-Efficient Deployment and Offloading Strategy in a Multi-AAV-Assisted MEC System

The mobile edge computing (MEC) system assisted by the unmanned aerial vehicle (AAV) is a promising technology to provide additional computing capability for mobile intelligent terminals (MITs). This article focuses on optimizing the energy consumption of a multi-AAV-assisted MEC system that serves a large number of MITs. To achieve this, a two-layer backtracking search algorithm (TBSA) is proposed. Specifically, the upper layer of TBSA aims to optimize the multi-AAV deployment by combining the backtracking search algorithm (BSA) with an adaptive population adjustment strategy based on generalized opposition-based learning. The lower layer of TBSA aims to determine the offloading decision and resource allocation based on the deployment of AAVs obtained from the upper layer algorithm. In the lower layer, a random priority sequence (RPS) is defined to describe the offloading decision of MITs and the BSA is employed to search for the optimal RPS. Simulation results have demonstrated the superiority of TBSA in the considered system, and its application scenarios are also discussed.