Rate-Splitting Multiple Access for Semi-Grant-Free Transmissions With MMD-Aided Grant-Based Users.

The increasing number of connected devices in machine-type applications, such as the Internet of Things (IoT), leads to the coexistence of heterogeneous services in Beyond 5G (B5G) networks, where users with diverse requirements share physical resources. In this paper, we apply a channel allocation method to assign Grant-Based (GB) users to channels. This method, named Max-Matching Diversity (MMD), leads to a frequency diversity gain that equals the number of available channels, reducing the outage of GB users. In the same network resource that serves a GB user, one Grant-Free (GF) user is allocated through a distributed contention protocol that considers the Quality-of-Service (QoS) of the GB user, performing the pairing that causes the minimum interference, so that the presence of a GF user is transparent to the GB user. Both users admitted to the same physical resource operate in either Non-Orthogonal Multiple Access (NOMA) or Rate-Splitting Multiple Access (RSMA), depending on the channel gains. Exact expressions for the outage probability of this Semi-Grant-Free (SGF) system are provided and compared to simulation results, showing that the SGF transmissions with MMD-aided GB users strategy reduces the outage and increases the achievable rate of both users.