Degree Type

Thesis

Date of Award

2010

Degree Name

Master of Science

Department

Computer Science

First Advisor

Johnny S. Wong

Abstract

New security threats emerge against mobile devices as the devices' computing power and storage capabilities evolve. Preventive mechanisms like authentication, encryption alone are not sufficient to provide adequate security for a system. There is a definite need for Intrusion detection systems that will improve security and use fewer resources on the mobile phone. In this work we proposed an intrusion detection method that efficiently detects intrusions in mobile phones using Data Mining techniques. We used network based approach that will remove the overhead processing from the mobile phones. A neural network classifier will be built and trained for each user based on his call logs .An application that runs on smart phone of the user collects certain information of the user and sends them over to the remote server. These logs then fed to the already trained classifier which analyzes the logs and sends back the feedback to the smart phones whenever abnormalities are found. Also we compared different neural classifiers to identify the classifier with better performance. Our results showed clearly the effectiveness of our method to detect intrusions and outperformed existing Intrusion detection methods with 95% detection rate.

Copyright Owner

Bharat Kumar Addagada

Language

en

Date Available

2012-04-30

File Format

application/pdf

File Size

30 pages

Share

COinS