Stochastic Lagrangian Krill Herd Optimized Quadratic Associative Boost Classification for Load Balancing in VANET

VANETs are wireless technologies specifically designed for communication among vehicles and roadside infrastructure. VANETs face various challenges, including network connectivity issues due to high mobility and limited communication range, especially in dense urban environments. Load balancing in VANETs is crucial for ensuring efficient and reliable data communication among vehicles. The dynamic nature of VANETs, characterized by rapidly changing network topologies and varying traffic loads, poses unique challenges for achieving optimal communication performance. In this paper, a Stochastic Lagrangian Krill Herd Optimized Quadratic Associative Boost Classification (SLKHO-QABC) method is introduced for resource-efficient load balancing in VANETs. The main objective of the proposed method is designed for load-balanced data transmission in VANETs with minimum end-to-end delay. The SLKHO-QABC method includes two major processes namely resource optimization and classification in VANET. Initially, the Stochastic Universal Sampled Lagrangian Krill Herd Optimization is used to determine the resource-efficient vehicle nodes based on fitness functions. With the optimal vehicle nodes, the load capacity is identified through classification. Quadratic Associative Boost Classification is utilized to categorize the less or heavy-loaded vehicle nodes based on the likelihood ratio test. Finally, the vehicle node with a higher load broadcasts an information packet to the lesser-loaded vehicle node during the time of flight, which is used to achieve efficient load balancing in VANET. Experimental analysis is performed for various parameters. Performance comparison analyses show that the proposed SLKHO-QABC method improves the load balancing efficiency throughput and minimizes energy utilization, packet loss rate, and end-to-end delay. SLKHO-QABC method improves the load balancing efficiency by 5.5.% and throughput by 40%, reduces the Packet loss rate by 31%, energy consumption by 18%, and end-to-end delay by 18.5%.