New framework for securing mobile adhoc network using lightweight authentication and signature-based intrusion detection system

• Main Author: Mandala, Satria
• Format: Thesis
• Language: English
• Published: 2012
• Subjects: T Technology (General)
• Online Access: http://eprints.utm.my/id/eprint/32318/1/SatriaMandalaPFSKSM2012.pdf; http://eprints.utm.my/id/eprint/32318/; http://dms.library.utm.my:8080/vital/access/manager/Repository?query=New+framework+for+securing+mobile+adhoc+network+using+lightweight+authentication+and+signature-based+intrusion+detection+system&queryType=vitalDismax&public=true
• Institution: Universiti Teknologi Malaysia

Mobile Adhoc Network (MANET) is vulnerable to network attacks due toits open communication medium. Blackhole and wormhole attacks are the mostsevere attacks in the network. The attacks cause congestion and increase thepossibility of confidential data theft. Unfortunately, the existing security solutionsare insufficient to protect the network. This work proposed a new securityframework, named Extra Secure Adhoc on Demand Distance Vector (ESAODV).This framework provides a defense-in-depth protection through layered securitymeasures: secure protocol and intrusion detection system (IDS) with extracountermeasures. The first layer implements lightweight packet authentication,and the second layer monitors and counters malicious packets. In this study,ESAODV was implemented using Java in Time Simulator/Scalable WirelessAdhoc Network Simulator, and analyzed using R-Statistics, Sigma Plot andMinitab. Results showed that ESAODV had contained the blackhole attackand the hybrid blackhole attack (HBHA) effectively. The number of corruptingrouting tables of benign nodes could be minimized to be near zero even if thenumber of attackers were increased. In addition, the IDS accurately detectedthe wormhole and the variant of wormhole attack called diversion of packet overthe wormhole link (DP-WHL). The false positive for live attack detection wassmall. The accuracy of detection was more than 94.5 percent. Although attackerschanged the pattern of packets diversion, the IDS detected the new attack patternin near real time. In addition to these findings, this research has also modeledfour performance metrics data of ESAODV, i.e., memory usage, elapsed timefor completing routing tasks, number of route replies and route success, basedon both linear regression and neural network. Goodness of fit parameters forthe models based on the neural network was higher than the linear regression. ESAODV has been proven to provide a comprehensive protection from the mostsevere attacks in the network. Furthermore, the performance metrics of ESAODVbased on the neural network produced a superior model.