Multi-Constraints Clustering Driven QoS-Centric VANET Routing Protocol: MCCQVR

In the last few years, Vehicle Ad-Hoc Network (VANET) has emerged as a potential wireless technology to serve different communication purposes including intelligent transportation, vehicle-to-infrastructure (V2I) and vehicle-tovehicle (V2V), internet-of-vehicular-things (IOVT) etc. Despite significances, the characteristics like high mobility, topologicaldynamism, link-vulnerability, iterative congestion make routing more challenging, especially in urban ecosystem. The existing routing protocols use standalone node parameter to perform routing decision; yet, its efficacy over dense deployed IOVT yields compromised performance due to the iterative linkoutage, retransmission cost and delay. Consequently, it impacts Quality-of-Service (QoS) aspects. Cluster-based routing protocols have performed better in densely deployed VOIT; however, ensuring stable clustering, optimal cluster-head (CH) selection and best-forwarding path formation remains the key to success. Ironically, the state-of-arts being developed over standalone feature driven solution could not meet IOVT demands. With this inference, this research paper proposes a robust multi-constraint (multi-metric) clustering-based QoScentric VANET routing protocol (MCCQVR) for IOVT communication. The MCCQVR protocol makes use of the multiple cross-layer parameters including node’s topology, packet velocity, link quality, and congestion information to perform CH selection. Additionally, it contributes service differentiation and adaptive resource allocation (SDARA) to guarantee optimally sufficient resource for real-time-traffic (RTD) transmission. Being a cross-layer protocol, MCCQVR exploits traffic details from the application layer, packet velocity or injection rate and congestion probability from the medium access control (MAC) layer, dynamic link quality from the datalink layer and neighborhood information from the network layer to perform CH selection followed by the best forwarding path estimation, which cumulatively guarantees transmission reliability over IOVT conditions. The SDARA on the other hand applied dual-buffer concept to retain reliable RTD transmission while guaranteeing optimally large resource for the non-realtime traffic (NRT). The simulation results over the different network conditions like payload, density, velocity etc. revealed that the proposed MCCQVR model achieved average PDR of 96.55% and 96% for the RTD and NRT traffic, respectively over the different payloads, speed and network density.