Unsupervised Clustering Algorithms for Reliable Routing and Data Dissemination using Optimization Techniques in Vehicular Adhoc Networks
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Today the usage of vehicles have increased tremendously. VANET is Vehicular Ad
newlinehoc Network which promotes the road safety among the vehicles. In recent years, VANETs
newlinehave emerged as a significant research area in the transportation sector. VANETs are
newlineconsidered as reliable networks for vehicular communication on highways or urban
newlineenvironments through the wireless medium. The purpose of VANET is to foster the QoS,
newlinehigh connectivity, bandwidth and security besides the number of challenges. The issues of
newlineVANET are routing and data dissemination due to its dynamic movement. Therefore, the
newlineresearch work tries to find best clustering algorithms and optimization techniques for
newlinereliable routing and data dissemination.
newlineThe purpose of this research is to form cluster with a selected CH (Cluster Head) and
newlineto select the optimal number of CHs. The path can be constructed from the source CH to the
newlinedestination CH by sending the beacon messages. The proposed research work are as follows
newlineMMBCMDSOR (Mean shift Margin Boost Clustering based Multivariate Dolphin Swarm
newlineOptimized Routing), BEXCMWO (Bagging Ensemble X-Means Clustering based Multi
newlinecriteria Whale Optimization), MIBCMCSO (Mutual Informated Brown Agglomerative
newlineClustering based Multiattribute Cockroach Swarm Optimization) and DLRSCODOR (Deep
newlineLearned Ruzicka Spectral Clustering based Oppositional Dragonfly Optimized Routing).
newlineThese research contributions were developed using the unsupervised clustering algorithms
newlineand optimization techniques to achieve routing and dissemination in an efficient manner. The
newlinemobility parameters such as density, direction, distance and velocity were used for
newlinepartitioning the network into clusters. To select the CH for a cluster, unsupervised clustering
newlinealgorithms were used and for the selection of optimal CHs various optimization techniques
newlinewere utilized. A beacon message is sent from the selected CHs to destination for the optimal
newlinepath selection and thus achieves data dissemination.
newlineAll the four proposed algorithms are impl