A Deep Reinforcement Learning-based resource allocation mechanism for XR applications*

Recently, the attention of academia and industry has turned towards 6G technology. Within 6G, eXtended Reality (XR) has emerged as an important scenario that is attracting researchers due to its requirements for computation-intensive, latency-sensitive, and high-bandwidth applications. Efficient resource allocation in XR is crucial for delivering a seamless and smooth user experience. However, existing research on resource allocation for XR applications does not consider the interaction between XR content and users. In this paper, we propose a deep Q network(DQN) approach to minimize service latency in XR applications. We formulate the problem as a Markov Decision Process(MDP) and apply DQN as the solution. Simulation results demonstrate the efficiency of the proposed scheme in terms of both latency and acceptance ratio.