Cost Minimization Strategy for UAVs Offload Calculation in MEC Environment

Abstract

To meet the cost minimization requirement of computational offloading for UAVs in Mobile Edge Computing(MEC) environment, this paper proposes a cost minimization strategy based on the improved DDQN optimization algorithm by time delays, energy consumption and computational offloading model. Aiming at the difficult problem of MEC server resource allocation in the model, this paper uses a dichotomous approximation solution strategy for optimal resource allocation, based on which an action screening strategy is adopted to avoid the dimensional disaster problem that may occur in the state space of DDQN, and finally priority is introduced on the empirical replay pool sampling, which is used to improve the convergence speed of the algorithm. The simulation experimental results show that the algorithm effectively reduces the overall system energy consumption and time delays, improves the task offloading success rate, and achieves good stability compared with several other classical algorithms.