Dr.Raheemaa Khan - Academia.edu (original) (raw)

Papers by Dr.Raheemaa Khan

Multi agent based intrusion detection architecture for the IDS adaptation over time

2017 Second International Conference on Electrical, Computer and Communication Technologies (ICECCT), 2017

Network security is an ever growing field. As, the information technology have conquered almost a... more Network security is an ever growing field. As, the information technology have conquered almost all domains, the search for loop holes in the security increased tremendously. This led to the deployment of various hardware and software based security tools. One of the tools used for implementing security is intrusion detection systems (IDS). We address a problem that current intrusion detection systems suffer. Intrusion detection systems highly rely on the previous patterns of attacks as well as the deviations of the normal patterns. This will lead to inefficiency as novel attacks can occur in the future due to the ever changing network and hosts configurations and technologies. This will make the IDS obsolete over time. Hence, adaptability of IDS to the future is a big issue. Here, we propose an agent based architecture that can adapt the IDS to the future threats.

In this busy world, information in the web is growing tremendously as the volume of data is updat... more In this busy world, information in the web is growing tremendously as the volume of data is updated daily on web. And most of the people use the internet search engine to find and retrieve the information. Since the data is more, retrieving the relevant document becomes difficult task. So this problem has paved a way to develop an algorithm on web content mining. In this research, Kendall’s Tau correlation analysis has been employed to determine the level liaison between the documents. The duplicate documents which are termed as outliers are identified if the correlation value between the documents is 1 and thus it can be eliminated. This method enforces the term frequency computation for common terms between the documents for which the ranking is done individually. From the experimental analysis, the proposed method provides better accuracy in detecting outliers in comparison with the existing methods. Keyword: Correlation, Web Content Mining, Correlation Coefficient, Ranking, Outl...

International Journal of Electrical and Computer Engineering (IJECE), 2019

Intrusion detection systems are used for monitoring the network data, analyze them and find the i... more Intrusion detection systems are used for monitoring the network data, analyze them and find the intrusions if any. The major issues with these systems are the time taken for analysis, transfer of bulk data from one part of the network to another, high false positives and adaptability to the future threats. These issues are addressed here by devising a framework for intrusion detection. Here, various types of co-operating agents are distributed in the network for monitoring, analyzing, detecting and reporting. Analysis and detection agents are the mobile agents which are the primary detection modules for detecting intrusions. Their mobility eliminates the transfer of bulk data for processing. An algorithm named territory is proposed to avoid interference of one analysis agent with another one. A communication layout of the analysis and detection module with other modules is depicted. The inter-agent communication reduces the false positives significantly. It also facilitates the iden...

Multi agent based intrusion detection architecture for the IDS adaptation over time

2017 Second International Conference on Electrical, Computer and Communication Technologies (ICECCT), 2017

Network security is an ever growing field. As, the information technology have conquered almost a... more Network security is an ever growing field. As, the information technology have conquered almost all domains, the search for loop holes in the security increased tremendously. This led to the deployment of various hardware and software based security tools. One of the tools used for implementing security is intrusion detection systems (IDS). We address a problem that current intrusion detection systems suffer. Intrusion detection systems highly rely on the previous patterns of attacks as well as the deviations of the normal patterns. This will lead to inefficiency as novel attacks can occur in the future due to the ever changing network and hosts configurations and technologies. This will make the IDS obsolete over time. Hence, adaptability of IDS to the future is a big issue. Here, we propose an agent based architecture that can adapt the IDS to the future threats.

In this busy world, information in the web is growing tremendously as the volume of data is updat... more In this busy world, information in the web is growing tremendously as the volume of data is updated daily on web. And most of the people use the internet search engine to find and retrieve the information. Since the data is more, retrieving the relevant document becomes difficult task. So this problem has paved a way to develop an algorithm on web content mining. In this research, Kendall’s Tau correlation analysis has been employed to determine the level liaison between the documents. The duplicate documents which are termed as outliers are identified if the correlation value between the documents is 1 and thus it can be eliminated. This method enforces the term frequency computation for common terms between the documents for which the ranking is done individually. From the experimental analysis, the proposed method provides better accuracy in detecting outliers in comparison with the existing methods. Keyword: Correlation, Web Content Mining, Correlation Coefficient, Ranking, Outl...

International Journal of Electrical and Computer Engineering (IJECE), 2019

Intrusion detection systems are used for monitoring the network data, analyze them and find the i... more Intrusion detection systems are used for monitoring the network data, analyze them and find the intrusions if any. The major issues with these systems are the time taken for analysis, transfer of bulk data from one part of the network to another, high false positives and adaptability to the future threats. These issues are addressed here by devising a framework for intrusion detection. Here, various types of co-operating agents are distributed in the network for monitoring, analyzing, detecting and reporting. Analysis and detection agents are the mobile agents which are the primary detection modules for detecting intrusions. Their mobility eliminates the transfer of bulk data for processing. An algorithm named territory is proposed to avoid interference of one analysis agent with another one. A communication layout of the analysis and detection module with other modules is depicted. The inter-agent communication reduces the false positives significantly. It also facilitates the iden...