Synergistic Superiorities of Employing IRS and RSMA in Downlink Cell-Free Massive MIMO Systems Under Finite Blocklength Regime

We explore the synergistic advantages of integrating intelligent reflecting surface (IRS) with rate splitting multiple access (RSMA) in a downlink cell-free massive multiple-input multiple-output (MIMO) system to meet the stringent requirements of ultra-reliable and low-latency communications. Taking into account the estimation errors, statistical channel knowledge, finite blocklength, and spatial correlation among IRS elements, a tight closed-form expression for the achievable rate is derived, which serves as a tool for evaluating the achievable rate across various system configurations. To enhance the weighted sum-rate (WSR) while adhering to the latency and reliability constraints, we formulate a joint WSR maximization problem with respect to both IRS phase shifts and power control coefficients. Given the non-convex nature of this problem, we develop an alternating optimization strategy that decouples the original problem into two distinct sub-problems. Specifically, the IRS phase shift design is reformulated as a min-max normalized mean squared error problem, enabling an efficient closed-form solution, whereas the power control optimization is addressed using a geometric programming approach. Numerical results validate the synergistic gain of integrating IRS with RSMA in terms of achievable rate and demonstrate that the proposed optimization scheme significantly enhances the WSR while fulfilling the latency and reliability requirements.