Archives

Hybrid Optimization and Support Vector Machine (SVM) Based Network Selection for Seamless Vertical Handover in Vehicular Networks (VANET)


S. Angelin Sophy and Dr.I. Laurence Aroquiaraj
Abstract

In the area of research of vehicular networks, Seamless mobility is a great dispute, which supports different applications dealing with the intelligent transportation system (ITS). Due to the vehicles speed and ever changing topology of the network, VANET continuous connectivity is a great dispute. So the selection of best available network becomes very difficult task. Vertical handover decision algorithm is proposed in this work, which chooses the best available network by comparing multiple parameter values (such as network traffic and vehicle speed). For vertical handover decision, Support Vector Machine (SVM) classification is chosen for selecting the best available network. The proposed algorithm is a hybrid model which merges the Biogeography-Based Optimization (BBO) and ant colony optimization. Initial one is utilized to allot the weight values to each service based on the quality of the services parameters to each sensor, where the weight values of each service is assigned according to the swarm intelligence. SVM classification method is applied, according to the weight value, which chose the specific node for vertical handover decision. By assessing the performance through the number of handoff occurs, throughput gained and level of latency attained, the algorithm is calculated.

Volume 11 | 08-Special Issue

Pages: 1476-1491