COINS: Towards a Correlation Based Intrusion Detection System for Mobile Ad hoc Network

Abstract

Mobile Ad-hoc Networks (MANET) encounters significant amount of security challenges due to their peculiar properties. To address these security challenges, this paper presents an approach towards secure data management in MANET. It presents an ingenious strategy based on the relationships among data items discovered via association rules mining algorithm. The proposed strategy assumes that the requests propounded by a particular host in a session are dependent on their underlying relationships and digression from these relationships should be regarded as abnormalcy. Based on this theory, a Correlation Based Intrusion Detection System (COINS) for MANET has been presented. The proposed solution considers the associations among requests of a session to establish the integrity of session and consequently of the data consumers. An analysis of the proposed scheme has been done under different scenarios that signifies the effectiveness of the proposed scheme.