Resource Allocation and Cooperation Scheduling for Reliability Maximization in Cluster Head Based Cooperative URLLC Networks

In this work, we consider a clustered ultra reliable low latency communication (URLLC) network, where one of the clustered users is identified as the cluster head for cooperatively improving the transmission reliability. In particular, we assume the access point (AP) performs no user scheduling and directly broadcast out the data for all the clustered users, in order to reduce the scheduling overload at AP. After the reception of the broadcast signal, each clustered user will attempt to decode the integrated large data packet and extract out desired data. To compensate the potential decoding error within the finite blocklength (FBL) regime, cooperative retransmissions will be scheduled from the cluster head to other users. Based on the practically restricted energy and blocklength resources, we formulate a maximum error probability minimization problem via jointly optimizing the resource allocation and cooperation scheduling from the cluster head. To address the nonconvex problem, we establish a tight convex approximation for the problem and propose an iterative algorithm for iteratively improving the solution until a convergence to a suboptimal point. Via simulations, we perform the algorithm validation, as well as the reliability evaluation, for our proposed solution. The performance advantages of our proposed solution and the impact of cluster head selection have also been illustrated.