Cost-Aware Task Offloading and Migration for Wireless Virtual Reality Using Interactive A3C Approach

Wireless virtual reality (VR) is becoming a promising service to provide users with immersive experience from anywhere. To deal with the performance-cost negotiation for a MEC-enabled wireless VR, we propose a cost-aware task offloading and migration scheme. We formulate the viewport rendering offloading, task migration, and subchannel allocation as an optimization problem, taking into account a long-term MEC operational cost budget and fluctuating channel conditions. To solve the problem, the Lyapunov optimization method is used to transform the long-term optimization to be a real-time optimization problem. Additionally, we design an interactive Asynchronous Advantage Actor-Critic (IA3C) algorithm to solve the problem in an online fashion. Our results show that the proposed scheme and the IA3C algorithm can substantially reduce the computing load of VR terminals while maintaining a low MEC operational cost in a high convergence rate.