An enhanced energy and distance based optimized clustering and dynamic adaptive cluster-based routing in software defined vehicular network

Abstract

Software-Defined Vehicular Networks (SDVN) have been established to facilitate secure and adaptable vehicle communication within the dynamic environment of Vehicular Ad-hoc Networks (VANETs). To enhance efficiency, various optimization techniques are employed in cluster-based routing, focusing on reducing energy consumption, improving cluster stability, enhancing throughput, minimizing network overhead, increasing packet delivery ratio, and reducing latency. This work proposes enhancements to dynamic adaptive cluster-based routing to mitigate suboptimal decisions in VANETs. A centralized controller maintains Energy and Distance-Based Clustering and Dynamic Adaptive Cluster-Based Routing (EDBC-DACBR) to optimize VANET clustering and routing. EDBC utilizes energy and distance metrics between vehicles and cluster centres, or Roadside Units (RSUs), for cluster formation. A fitness model identifies Cluster Heads (CH) based on nodes with the highest fitness values, while a Location-Based Fuzzy C-Means (LBFCM) algorithm ensures optimal cluster formation. The resultant CH, chosen for their energy efficiency, stability, and dynamism, are derived by combining the LBFCM with the fitness model. Additionally, DACBR adapts to network variations, such as energy levels, communication distances, and vehicular congestion, to define the shortest path. Simulation-based evaluations demonstrate the effectiveness of the proposed approach, outperforming existing methods such as Learning-Based Cluster-Based Routing (ANFC-QGSOR), Fuzzy-Based Cluster-Based Routing (FCBR), Energy-Efficient-Based Cluster-Based Routing (EEOR), and Hierarchy-Based Cluster-Based Routing (EHCP) in terms of throughput, overhead, packet loss, latency, stability, and network lifetime. Specifically, EDACR achieves a 15% improvement in throughput, reduces network overhead by 20%, increases the packet delivery ratio by 25%, and decreases latency by 30% compared to existing approaches. Furthermore, EDACR enhances network stability, with a 10% reduction in packet loss and a 20% increase in network lifetime. These results highlight the efficacy of EDACR in enhancing the efficiency and reliability of SDVN deployments in dynamic vehicular environments.