Multi-Source Multi-Destination Hybrid Infrastructure-Aided Traffic Aware Routing in V2V/I Networks

The concept of the “connected car” offers potential for safer, more enjoyable and efficient driving eventually autonomous driving. However, in urban Vehicular Networks (VNs), high mobility vehicles along roads poses major challenges to routing protocols needed a reliable flexible vehicular communications system. Thus, VNs rely on static Road-Side-Units (RSUs) forward data extend coverage across network. In this paper, we first propose new Q-learning-based algorithm, namely Infrastructure-aided Traffic-Aware Routing (I-TAR), which leverages wired RSU infrastructure packet forwarding. Then, focus multi-source, multi-destination problem effect imposes node availability, as nodes also participate other paths. This motivates our hybrid approach, Hybrid Traffic Aware (HI-TAR) that aims select best Vehicle-to-Vehicle/Infrastructure (V2V/I) route. Our findings demonstrate I-TAR can achieve up 19% higher average packet-delivery-ratio (APDR) compared state-of-the-art. Under realistic scenario, where availability is considered, decline 51% APDR performance observed, whereas proposed HI-TAR turn increase by 50% both Finally, when multiple source-destination vehicle pairs are all schemes model consider i.e. limited-availability, from 72.2% 82.3% lower APDR, those do not, assuming full-availability. still provides 34.6% better than I-TAR, ~40%