A Joint Framework of HARQ and RSMA for AoI Minimization

The advent of sixth-generation (6G)-enabled real-time applications, such as remote surgery and autonomous driving, has sparked significant interest in a novel and integrated metric known as the Age of Information (AoI). This metric is particularly valuable because it effectively reflects the tradeoffs between reliability, latency, and frequent updates. In this article, we propose an intelligent framework for hybrid automatic repeat request (HARQ) enabled rate-splitting multiple access (RSMA) networks, aiming to optimize resource utilization while maintaining a balance between timeliness and reliability. Particularly, the source intelligently adapts power allocation and retransmissions for common and private streams of RSMA carrying data packets of different priorities, considering the current state of AoI and feedback signals. First, an analytical framework is established by deriving closed-form approximations for the individual average block error rates (BLERs) of both common and private streams. Subsequently, an AoI minimization problem is formulated using the derived BLER. Following that, the problem is modeled as a Markov decision process (MDP), allowing us to develop an optimal policy for minimizing the system's average AoI. Moreover, we also devise a near-optimal solution utilizing the Lyapunov drift function, accounting for the MDP's complexity. Extensive simulations demonstrate that the proposed approach improves the AoI performance of the RSMA scheme and outperforms the conventional approaches.