Task Offloading and Parameter Optimization of MAR in Multi-access Edge Computing

With the popularization of mobile devices and the development of mobile augmented reality (MAR), people's demand for MAR applications is increasing. However, few MAR applications are applied to mobile devices since it is computationally intensive and requires computing power. MAR applications consume a lot of energy, and the battery capacity of mobile devices cannot support the long-term operation of applications. In this paper, we study the problem of dynamically configuring parameters of MAR system based on different user service characteristics under the system model of multi-edge server and multi-client. Considering the influence of client parameters such as CPU frequency, client task resolution and wireless bandwidth resource allocation, we design an objective function to evaluate service delay, energy consumption per frame and detection accuracy of target tasks according to different user service characteristics. Based on our objective function, we propose a server selection and parameter optimization (SSPO) algorithm to select the appropriate mobile edge server for client task offloading and parameter optimization. Simulation results show that our proposed algorithm can effectively reduce the energy consumption of MAR client while meeting the service characteristics of different users.