Reliable Wireless Networking in Highly Dynamic Environments: Do Partial Link Statistics Suffice?

Numerous radio units (RUs) are required to densely compose small cells in the beyond 5G mobile networks. The concept of vehicle-mounted RUs is a promising solution for dynamically deploying small cells in accordance with demand distribution. Wireless relay fronthaul networking is the key enabler for this concept, where forwarding paths of fronthaul streams are computed in real-time by an edge server using the link-state information reported from RUs. Optimization of the reporting interval is a significant issue because of the tradeoff between the freshness of report messages and network load. However, existing works have not investigated the freshness of link information in highly dynamic environments. In this paper, we introduce a new reliability metric called the reliability of information (RoI). The RoI is defined as the joint probability that all nodes in the network maintain correct information. We establish a tractable lower bound of the RoI under a mild assumption on the dynamics of node connectivity, which enables us to design small cells guaranteeing information reliability. We further formulate and solve an optimization problem to find an optimal reporting interval. The usefulness of the proposed scheme is demonstrated through simulation experiments for a highly dynamic vehicular network.