Performance Evaluation of TGSR-DSDV Routing Under Varying Traffic Conditions in VANETs
Abstract
Dynamic traffic conditions represent one of the most disruptive factors in Vehicular Ad Hoc Network (VANET) routing, inducing network congestion, elevated packet drop, and increased delay. Conventional proactive routing protocols such as Destination-Sequenced Distance Vector (DSDV), with their static parameter configurations, are inherently incapable of adapting to network topology fluctuations in traffic load. This paper evaluates the performance of TGSR-DSDV, a bio-inspired enhanced routing protocol that integrates The Great Salmon Run (TGSR) optimization into the DSDV protocol framework to optimize routing parameters helloInterval, routeLifetime, and broadcastDelay according to scenario-specific VANET conditions. Multi-scale performance is evaluated across three urban VANET environments Pauh (small), Arau (medium), and Alor Setar (large) under varying packet sizes (64B to 1024B) using a User Datagram Protocol (UDP) burst traffic model. Simulations are evaluated in OMNeT++ 5.6.2 with INET 4.2.5 framework and IEEE 802.11p MAC. The proposed protocol is evaluated in terms of Packet Delivery Ratio (PDR), throughput, and end-to-end delay under heterogeneous traffic generation conditions. Additional cross-protocol comparisons with DSDV, Ad hoc On-Demand Distance Vector (AODV), Dynamic MANET On-demand (DYMO), and Greedy Perimeter Stateless Routing (GPSR) are conducted under identical simulation settings to validate the effectiveness of TGSR-DSDV. Experimental results demonstrate that TGSR-DSDV consistently improves routing performance and traffic adaptability across varying packet sizes and VANET scales by achieving an average PDR of 94.56%, average throughput of 458.50kbps, and an average end-to-end delay of 0.0174 s in both cross-protocol and multi-scale evaluation. The findings highlight the capability of TGSR-DSDV to provide improved traffic adaptability, routing reliability, and communication efficiency for urban VANET applications.