Vehicular massive multiway relay networks applying graph-based random access

Recent vehicular communications problems entail various aspects, such as rapidly changing topologies and large number of users. We consider massive uncoordinated multiway relay networks (mu-mRN) that can serve a massive number of users expecting to fully exchange information among them via a common multiway relay. Relay vehicular networks system is one of potential applications of the mu-mRN. In this paper, we are interested in improving normalized throughput T of the network using multiuser detection (MUD) capability of K > 1 based on the graph-based random access to flexibly adapt the topology changes. First, we present a network capacity bound of the mu-mRN with general K to investigate the theoretical limit of the networks. Then, we search for many optimal and practical degree distributions for each theoretical bound. Second, we aim to improve the normalized throughput by 10× from the maximum normalized throughput of conventional systems. To achieve the goal, we propose the mu-mRN applying doubly irregular coded slotted ALOHA. We also propose an optimal practical encoding of the mu-mRN to closely approach the target with finite number of users.