Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/459053
Title: | Performance enhancement of intrusion detection system using dimensionality reduction techniques and evaluation with different machine learning classifiers on optimal dataset |
Researcher: | Surya Prakash J |
Guide(s): | Suguna R |
Keywords: | Traffic Classification Machine Learning Classifiers Intrusion Detection Systems |
University: | Anna University |
Completed Date: | 2022 |
Abstract: | Traffic classification is an automated process which categorizes computer network traffic based on various parameters such as port number or protocol. Traffic classification is an essential tool for network and system security in complex environment. Intrusion detection is a monitoring system that detects suspicious activities and generates alerts. Network Intrusion Detection Systems (NIDS) play an important role to monitor and analyze network traffic to protect a system from network-based threats. The Intrusion Detection Systems (IDS) are of different types - Active and passive IDS, Network Intrusion Detection Systems (NIDS), Host Intrusion Detection Systems (HIDS), Knowledge-based (Signature-based) IDS and behavior-based (Anomaly-based) IDS. The Active IDS is also known as Intrusion Detection and Prevention System and Passive IDS is configured to only monitor and analyze network traffic activity and alert an operator to potential vulnerabilities and attacks. newlineA Network-based Intrusion Detection System (NIDS) detects malicious traffic on a network. Host-based IDS runs on a host and monitors system activities for signs of suspicious behavior. Signature-based detection is typically best used for identifying known threats. Anomaly-based intrusion detection systems can alert the suspicious behavior that is unknown. Network Traffic datasets are captured from real time network using packet sniffer and analysis tool. The intrusion detection system developed based on flow and payload statistical features with clustering technique requires more number of clusters for un-identified traffic network. Also it is difficult to map large number of clusters to small number of real time applications. Though this method is more effective, the design process is more complex. The research requires suitable feature selection algorithms and optimal dataset to enhance the accuracy. newline |
Pagination: | xviii,110p. |
URI: | http://hdl.handle.net/10603/459053 |
Appears in Departments: | Faculty of Information and Communication Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 120.32 kB | Adobe PDF | View/Open |
02_prelim pages.pdf | 2.64 MB | Adobe PDF | View/Open | |
03_content.pdf | 32.11 kB | Adobe PDF | View/Open | |
04_abstract.pdf | 9.67 kB | Adobe PDF | View/Open | |
05_chapter 1.pdf | 311.91 kB | Adobe PDF | View/Open | |
06_chapter 2.pdf | 412.49 kB | Adobe PDF | View/Open | |
07_chapter 3.pdf | 493.78 kB | Adobe PDF | View/Open | |
08_chapter 4.pdf | 591.21 kB | Adobe PDF | View/Open | |
09_chapter 5.pdf | 678.73 kB | Adobe PDF | View/Open | |
10_chapter 6.pdf | 453.98 kB | Adobe PDF | View/Open | |
12_annexures.pdf | 116.09 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 126.73 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: