QoE-Driven Computation Offloading: Performance Analysis and Adaptive Method
Quyuan Luo, Weisong Shi, Pingzhi Fan · 2021 13th International Conference on Wireless Communications and Signal Processing (WCSP) · 2021
With the proliferation of the Internet of Things (IoT) and the concomitant computation-intensive tasks, the surging demand for computation offloading can be expected. How to optimally make offloading decisions towards best quality of experience (QoE) represents a fundamental research problem. In this paper, considering the different QoE requirements and the inherent characteristic of tasks, we investigate the computation offloading problem in edge computing enabled IoT networks. Particularly, we first establish task local computing model and task offloading model by fully considering the parallel and serial processing characteristics of tasks. Based on a thorough theoretical performance analysis, a QoE-driven adaptive computation offloading (QEACO) strategy is proposed. And users can optimally and adaptively make offloading decisions towards best QoE. Finally, simulation results indicates that QEACO can significantly improve the QoE of users compared to several benchmark schemes.