Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/22726
Title: | Intelligent intrusion detection techniques for mobile Ad Hoc Networks |
Researcher: | Ganapathy, S |
Guide(s): | Kannan, A |
Keywords: | algorithms Detection Techniques false alarm IAASA Intrusion Mobile Ad Hoc Networks |
Upload Date: | 11-Aug-2014 |
University: | Anna University |
Completed Date: | n.d. |
Abstract: | In this thesis a new intelligent intrusion detection system has been newlineproposed and implemented to reduce the false alarm rate present in the newlineexisting intrusion detection systems For this purpose two new feature newlineselection algorithms namely an intelligent agent based attribute selection newlinealgorithm and an intelligent CRF based feature selection algorithm have been newlineproposed These algorithms are capable of selecting the most optimal number newlineof features from the data set This helps to reduce the classification time so newlinethat the detection algorithm can work in real time newlineIn this thesis two new feature selection algorithms have been newlineproposed and implemented for selecting the optimal number of features to newlineimprove the classification accuracy The Intelligent Agent based Attribute newlineSelection Algorithm IAASA has been proposed in this work to improve the newlineclassification performance by selecting the features which will contribute newlinesignificantly in the classification process In this model the KDD cup data set newlinewhich consists of 41 features is considered as input to the algorithm since it is newlinea bench mark data set for intrusion detection The IAASA uses the newlineInformation Gain Ratio IGR value in order to select only the most important newlinefeatures from the data set Therefore the IGR values are computed for all the newline41 attributes which are present in the data set Moreover the proposed newlinealgorithm works on subsets of the data set and hence a major subset selected newlineat random from the KDD cup data set is split into number of subsets each newlinetime and the average IGR value for each subset is considered and compared newlinewith a threshold value to make a final decision In this way 19 features have newlinebeen identified by this process from the original 41 features Moreover new newlineintelligent agents have been proposed in this work for effective feature newlineselection so that the agents act intelligently to select the optimal number of newlinefeatures which are used for classifying the intruders based on their behaviors newline newline |
Pagination: | xix, 177p. |
URI: | http://hdl.handle.net/10603/22726 |
Appears in Departments: | Faculty of Information and Communication Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 15.93 kB | Adobe PDF | View/Open |
02_certificate.pdf | 800.26 kB | Adobe PDF | View/Open | |
03_abstract.pdf | 16.94 kB | Adobe PDF | View/Open | |
04_acknowledgement.pdf | 8.28 kB | Adobe PDF | View/Open | |
05_content.pdf | 41.95 kB | Adobe PDF | View/Open | |
06_chapter 1.pdf | 112.78 kB | Adobe PDF | View/Open | |
07_chapter 2.pdf | 136.71 kB | Adobe PDF | View/Open | |
08_chapter 3.pdf | 30.86 kB | Adobe PDF | View/Open | |
09_chapter 4.pdf | 308.81 kB | Adobe PDF | View/Open | |
10_chapter 5.pdf | 193.26 kB | Adobe PDF | View/Open | |
11_chapter 6.pdf | 732.9 kB | Adobe PDF | View/Open | |
12_chapter 7.pdf | 91.02 kB | Adobe PDF | View/Open | |
13_chapter 8.pdf | 20.89 kB | Adobe PDF | View/Open | |
14_reference.pdf | 792.86 kB | Adobe PDF | View/Open | |
15_publications.pdf | 51.08 kB | Adobe PDF | View/Open | |
16_vitae.pdf | 5.24 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: