Joint Source Rate Control and Radio Resource Allocation for QoE Maximization in RSMA Networks

The quality of experience (QoE) perceived by users in wireless data services is significantly influenced by source rates and the delays caused by queuing and transmission over wireless links—factors that have not been considered in resource allocation research for rate-splitting multiple access (RSMA). In this paper, we address this gap by formulating a QoE maximization problem for RSMA resource allocation, modeling QoE as a combination of logarithmic user satisfaction with source rates and delays, based on GI/G/1 queuing delays with general distributions of data interarrival times and sizes. We reveal that this problem can be decomposed into sum-log-rate maximization and link utilization optimization for individual users. To solve these subproblems, we develop an alternating optimization (AO) algorithm. The sum-log-rate maximization subproblem is further decomposed into precoder optimization—solved using the weighted minimum mean square error (WMMSE) method—and common-rate division, for which a closed-form solution is derived given a Lagrange multiplier obtained via bisection. For the link utilization optimization subproblem, we apply successive convex approximation (SCA) in the general GI/G/1 queuing case, while deriving a closed-form solution for the special case of an M/M/1 queuing system. The signaling overhead, computational complexity analysis, and convergence proof of the proposed solution algorithm are also provided. Our numerical evaluation demonstrates that it consistently outperforms the reinforcement learning-based, sum-rate maximization, and max-min fairness algorithms in terms of sum QoE. A notable finding from our evaluation is that the division of the common rate among users plays a crucial role in QoE maximization, which markedly differs from the outcomes of conventional approaches.