Construction of a stable vehicular ad hoc network based on hybrid genetic algorithm

Abstract

In vehicular ad hoc networks, the vehicle speed can exceed 120 kmph. Therefore, any node can enter or leave the network within a very short time. This mobility adversely affects the network connectivity and decreases the life time of all established links. To overcome these problems, many routing protocols based clustering technique have been proposed. Indeed, the poor assignment of vehicles to clusters is the most important shortcomings where an inaccurate affiliation may reduce the algorithm's effectiveness and disrupt all results and analyzes. Therefore, in this paper, we used a hybrid genetic algorithm to improve the cluster maintenance phase in our Weighted K-medoid Clustering Algorithm (WKCA) proposed recently. The proposed model incorporated the tabu search within genetic algorithm to allow the scan of all search space and to reach the best solution without falling into the local optima. This model improves the assignment of nodes to clusters, which in turn achieves efficient vehicle communication and ensures more stability in clustered architecture. Based on relevant metrics, the results obtained in a simulation game show that the enhanced model (E-WKCA) achieves more stability and robustness when compared to the original algorithm (WKCA) and other approaches designed for the same objective.

ABNC: Adaptive Border Node Clustering Using Genes Fusion Based on Genetic Algorithm to Support the Stability of Cluster in VANET

“…The Enhanced Model of Weighted K-Medoids Clustering Algorithm (E-WKCA) is the first paper that discusses border cluster issues. It was conducted by Hajlaoui in 2019 [4]. Their proposed method used the combination of metaheuristic algorithms.…”

“…In an adjacent clustering process, it will create a cluster border; as a consequence, the connection stability is interrupted. An evaluation of this cluster border has been investigated using a combination of the Genetic Algorithm and Tabu Search methods [4]. These algorithms detected the similarity of vehicles in the border cluster area with the cluster head.…”

“…Each vehicle on the border cluster will have two choices: staying in the initial cluster or moving to a new cluster that matches the characteristics of the vehicle. The border cluster phenomenon becomes an interesting area to be resolved [4]. Though only a few researchers concern in this area, there is still room for improvement that can be extracted in this border cluster.…”

“…It is slightly different from the previous method based on the original Genetic Algorithm. In original Genetic Algorithm, the parent's and the children's genes will be combined into one chromosome [4]. They didn't use the temporary cluster to increase the stability of main cluster.…”

“…The throughput shows the stability under any condition. We compared the results of the study with the An Adjusted K-Medoids Clustering Algorithm for effective stability in vehicular ad hoc networks (AKCA) [7] and Enhanced Model of Weighted K-Medoids Clustering Algorithm (E-WKCA) [4] methods as the recent methods running in the cluster problems.…”