Stochastic Geometry Analysis of a Two-Tier Vehicular Network with Roadside Units and Vehicular Relays

This research explores the utilization of relays in vehicle-to-all (V2X) communications, where roadside units (RSUs) alone may not be sufficient to ensure network connectivity due to network congestion, signal attenuation, or interference. By employing stochastic geometry, we analyze a spatially-correlated vehicular network that incorporates both RSUs and relays to serve network users on roads. Our model considers the geometric characteristics of roads, RSUs, relays, and users using Cox point processes conditionally on the same road structure. Assuming disjoint frequency resources for serving users and for enabling RSU-connected relays, each user can associate with either an RSU or a relay whichever is closest to it. We derive the association probability and coverage probability for the typical user, enabling us to assess network performance. Additionally, we investigate user throughput by considering interactions among different links within the proposed network. This paper offers practical insights for the design of two-tier vehicular networks. Specifically, we express user association, user signal-to-interference ratio (SIR), and user throughput as functions of network variables. This information aids in determining optimal relay density and operating bandwidth to enhance network reliability and maximize user throughput in vehicular networks.