Dynamic Statistical Data Based Intrusion Detection Scheme in Manet using Fuzzy Rules

Abstract

The Mobile Adhoc Network (Manet) has been well analyzed and explored towards the problem of intrusion detection. To detect the intrusion attack, there exist numerous techniques to handle this issue; even though, they struggle to achieve expected performance. To improve the performance an efficient dynamic statistical data based intrusion detection scheme is presented in this paper.  The method handles the intrusion detection according to the statistics about the network which represent the various features like traffic, latency, and throughput, number of service access and so on. According to these statistics, the incoming packet has been analyzed to measure time domain legitimate score (TDLS). The TDLS measure represents the suitability of the incoming packet and represents the trustworthy also. The value of TDLS is measured based on the fuzzy rule available where there are number of rules being generated and maintained according to the statistics considered. Based on the statistic values a specific rule has been selected to measure TDLS value. According to the value of TDLS, the method performs intrusion detection and improves the performance also.