Investigation Leading to Behaviour-Based Hybrid Intrusion Detection System for Mobile Devices

Work in Progress
Khurram Majeed

Smartphones nowadays have become immensely popular because they provide All-In-One expediency by integrating traditional mobile phones with hand-held computing devices making them more open and general purpose. However this flexibility leaves the Smartphones prone to attacks by malicious hackers. These malware not only poses a threat to mobile system data confidentiality, availability and integrity but can result in unwanted billing, depletion of battery power and denial-of-service (hereafter DOS) attack by generating malicious traffic hence seriously crippling the mobile network and service capacity. Current Smartphones malware detection and prevention techniques are limited to Signature-Based antivirus scanners (hereafter SIDS). These can efficiently detect malware with a known signature, but they have serious shortcomings with new and unknown malware creating a window of opportunity for attackers. A framework for Behaviour-Based Hybrid Intrusion Detection System is proposed to circumvent these shortcomings. This framework aims to provide protection against physical misuse using Machine Learning technique and detection of malicious applications using Knowledge Based Temporal Abstraction method. This research will be among the first to combine these two methods. Being platform independent is another novelty of the framework. A prototype has been partially implemented on Google Android and tested on emulators. Further validation will be performed on Smartphones to benchmark this framework.

Type of Publication: Paper
Conference: PPIG Doctoral Consortium 2012
Publication Year: 2012
Paper #: 12
TitleInvestigation Leading to Behaviour-Based Hybrid Intrusion Detection System for Mobile Devices
Publication TypePaper
AuthorsMajeed, K
PPIG Workshop: 
2012-11-WIP