Relaying-Assisted Multiuser Networks in FBL Regime: Achievable Reliability-Constrained Throughput

In this work, we study a multi-user relaying network operating with finite blocklength (FBL) codes, where a single relay node is responsible for relaying the data from a source node to multiple users. We consider two relaying principles, i.e., decode-and-forward (DF) and amplify-and-forward (AF) relaying, together with two downlink strategies for relaying to multiple users, i.e., broadcasting and time division multiple access (TDMA), which result in total in four combinations. Taking the fairness into account, we formulate optimization problems to maximize the minimum throughput among all users via blocklength allocation, while guaranteeing the reliability. We start with the combination strategy of TDMA-DF, where the DF relay serves multiple users in a TDMA manner in the second hop. To solve the nonconvex problem, we have constructed a convex approximation for the problem and accordingly proposed iterative algorithm, which is capable of iteratively improving the minimum throughput until a convergence to a suboptimal point. Afterwards, the iterative algorithm is then extended to other combinations, i.e., broadcasting-DF and TDMA-AF, while for broadcasting-AF, the optimal blocklength allocation can be directly derived out. Finally, via simulation results, we validate the convergence of our proposed iterative algorithms, and by comparison depict the benefits of DF relaying over AF relaying and the advantages of broadcasting and TDMA strategies in different scenarios.