Data-Oriented Uplink RSMA Systems: Performance Analysis and Design Insight

Abstract

In this article, we study the timely notion of short-packet communications, with specific focus on uplink rate-splitting multiple access (RSMA) systems under the finite blocklength regime. Specifically, we consider rate-adaptive and power-adaptive uplink RSMA mechanisms, incorporating multiple user groups, and derive analytical expressions for the fundamental delay-outage rate (DOR) metric. The analytical derivations are validated through the corresponding Monte Carlo numerical simulations, reflecting high accuracy. Then, the derived DOR expressions are exploited for providing optimal DOR performance under diverse individual transmission parameters, such as different information delivery time thresholds, varying blocklengths and channel bandwidths, while also considering the non-orthogonal multiple access (NOMA) as a particular special case. Importantly, DOR optimization in terms of the message splitting ratio is also pursued, and a feasible optimization algorithm is proposed. A vast collection of numerical results is then provided, comparing between the rate-adaptive and the power-adaptive schemes, assessing the impact of message splitting ratio optimization while also comparing between RSMA and NOMA. Additionally, the overall power efficiency and impacts of imperfect channel state information (CSI) are assessed and shown. Overall, the offered analysis methods and numerical results provide valuable tools and insight for deploying, designing and optimizing data-oriented uplink RSMA mechanisms in future wireless systems such as the emerging 6G networks.