Research Article
Distributed Deep Reinforcement Learning Based Mode Selection and Resource Allocation for VR Transmission in Edge Networks
@INPROCEEDINGS{10.1007/978-3-030-99200-2_13, author={Jie Luo and Bei Liu and Hui Gao and Xin Su}, title={Distributed Deep Reinforcement Learning Based Mode Selection and Resource Allocation for VR Transmission in Edge Networks}, proceedings={Communications and Networking. 16th EAI International Conference, ChinaCom 2021, Virtual Event, November 21-22, 2021, Proceedings}, proceedings_a={CHINACOM}, year={2022}, month={4}, keywords={Virtual reality Reinforcement learning Resource allocation Mobile edge network}, doi={10.1007/978-3-030-99200-2_13} }- Jie Luo
Bei Liu
Hui Gao
Xin Su
Year: 2022
Distributed Deep Reinforcement Learning Based Mode Selection and Resource Allocation for VR Transmission in Edge Networks
CHINACOM
Springer
DOI: 10.1007/978-3-030-99200-2_13
Abstract
Wireless virtual reality (VR) is expected to be one of the most pivotal applications in 5G and beyond, which provides an immersive experience and will greatly renovate the way people communicate. However, the challenges of VR service transmission to provide high quality of experience (QoE) and a huge data rate remain unsolved. In this paper, we formulate an optimization of the mode selection and resource allocation to maximize the QoE of VR users, aiming at the optimal transmission of VR service based on the cloud-edge-end architecture. Moreover, a distributed game theory based deep reinforcement learning (DGTB-DRL) algorithm is proposed to solve the problem, which can achieve a Nash equilibrium (NE) rapidly. The simulation results demonstrate that the proposed method can achieve better performance in terms of training efficiency, QoE utility values.
