Computing Resource Allocation and Offloading Method based on Simulated Annealing Algorithm

Abstract

The integration of mobile edge computing (MEC) and 5G ultra-dense small cell network structure provides higher computing capability for smart mobile devices (SMDs). This paper proposes a joint optimization method for task offloading and computing resource allocation based on Simulated Annealing Algorithm (SAA) in small cell wireless network scenarios with multiple base stations (BSs) and multiple MEC servers. Firstly, this method jointly considers computing resource allocation and task offloading. Secondly, it uses the Lagrange multiplier method to solve the problem of computing resource allocation. Thirdly, it adjusts the offloading decision through the improved simulated annealing algorithm (improved-SAA) dynamically. This paper uses the task completion time and the total weighted energy consumption as the System Utility Function to measure the benefit of task offloading. Compared with hjTORA Algorithm, Greedy Algorithm, and Local Search Algorithm, the experimental results show that the algorithm has better utility value under the influencing factors such as the number of subcarriers and SMDs, communication data, calculation data and other factors, which verifies the effectiveness of the designed algorithm.