In 5G networks and beyond, enhanced Mobile Broadband (eMBB) service targets at providing extremely high throughput, Ultra-Reliable and Low Latency Communications (URLLC) service supports high reliability and low latency while massive machine-type communication (mMTC) aims at high connection density. Different types of traffic have different requirements, which brings difficultity to serve different users on the same resource block. In this paper, we investigate the co-existence of eMBB, URLLC and mMTC services based on superposition scheme in the 5G downlink scenario. To minimize the average data rate loss of eMBB users while satisfying the latency and reliability constraints of URLLC and mMTC users, we formulate a mixed-integer nonlinear program (MINLP) problem which is non-convex. We propose Successive Convex Approximation (SCA) Method and Proximal Policy Optimization (PPO), an advanced deep reinforcement learning (DRL) algorithm, to solve the non-convex problem. In the proposed algorithm, the agent seeks for the optimal user pairing and power allocation policy when the URLLC traffic arrives sporadically. Simulation results depicts the performance of our proposed PPO based superposition scheme nears to SCA-based method with lower complexity.
Joint Resource Allocation for Multiplexing eMBB, URLLC and mMTC Traffics Based on DRL
24.06.2024
906297 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Cooperative UAV-Assisted Joint Scheduling of Coexisting URLLC and eMBB Services
Springer Verlag | 2023
|