Hybrid Machine Learning Procedure to Handle Hybrid Attacks in Wireless Vehicular ad hoc Networks
Pavan Kumar B V S P1, S.S.V.N. Sarma2, C. Lokanatha Reddy3
1Pavan Kumar B V S P, Scholar, Professor, Department of Computer Science, Dravidian University, Kuppam Malla Reddy Engineering College for Women, Hyderabad (Telangana), India.
2Dr. S.S.V.N. Sarma, Dean, Vaagdevi Engineering College, Warangal (Telangana), India.
3Dr. C. Lokanatha Reddy, Dean, Department of Science & Technology, Dravidian University, Kuppam (Andhra Pradesh), India.
Manuscript received on 25 May 2019 | Revised Manuscript received on 03 June 2019 | Manuscript Published on 22 June 2019 | PP: 182-185 | Volume-8 Issue-3S, February 2019 | Retrieval Number: C10370283S19/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Vehicular ad hoc networks (VANETs) are an emerging technology in modern environment transportation media. Because of VANET importance in information circulating through network is a crucial life in real time scenario. Combining information extraction from different vehicles is a complex need for real time applications. The volatile nature of the communication connections in network has made up VANET vulnerable to different types of security related attacks. Sybil, Distributed Denial of Service (DDOS) attacks are the major attack sequences that exhausts the network by illegitimately based on its resources. In this type attack sequences different types of fake identifiers consists spoofed vehicle id’s based on related server ip_address to exhaust the network by circulating bogus messages from other vehicles present in VANETs. So that in this paper we propose Hybrid Machine Learning approach (which consists Support vector machine (SVM), artificial neural network classification and AODV protocol hierarchy) to handle Sybil with DDOS attack sequences and provide efficient communication between vehicle nodes in vehicular ad hoc networks. Attacks handle in this scenario consist two basic steps i.e first selects most relevant feature from network based on data transmission from one to different vehicular nodes, based on selected feature classify the attack and then handle the efficient data transmission process in vehicular ad hoc networks. Experimental results of proposed approach gives better simulation parameters with respect to delivery ratio, throughput and others in vehicular ad hoc networks.
Keywords: Vehicular Ad Hoc Networks, Denial Of Service Attacks, Sybil Attacks, Support Vector Machine, Network Communication, AODV.
Scope of the Article: Machine Learning