Analysis and Development of Intrusion Detection Techniques

Abstract

In this new era, information plays a key role for everyone, compromising with information newlinemay be harmful to the user or our society. Advancements in the network infrastructure have caused a positive influence in our day-to-day life. Many reform initiatives have been taken newlineall over the world which are related to the digitization of the country s methodologies of newlinehandling in-formation. The usage of modern techniques also has a drawback, which allows newlinedata theft. Hence, a secure system is required, which can detect any kind of fraudulent newlineactivity and alert the administrator. Intrusion detection is a very useful tool to protect the newlineinformation at host level as well as network level. An Intrusion Detection System (IDS) is newlinea network security system that detects, identifies, and tracks an intruder or an invader in a network. As the usage of the internet is growing every day in our society, the IDS is becoming an essential part of the network security system. Therefore, the proper research newlineand implementation of IDSs are required. With the help of improved technologies at our newlinedisposal, many solutions have been found to create many intrusion detection systems. newlineHowever, it is difficult to identify the perfect solution from the vast options we have available. Many researchers have proposed several techniques for Intrusion Detection. Presently, Hybrid Intrusion Detection is a prime area of research in the field of Intrusion Detection System. In Hybrid Intrusion Detection, more than one detection method are newlineconsidered together to take advantage of many techniques in a single model. Hybrid newlineschemes like Neural and Fuzzy Logic, Binary PSO and random forests algorithm, newlinecombining decision tree with Naive Bayes, Neural based Hybrid Technique, Neural newlinenetwork and Bayesian Networks, Neural with Genetic algorithms, HMM with Naïve newlineBayesian, RBF and Elman Neural Network, Self-Organizing Map and Back Propagation newlineNetwork, and many others which will be mentioned in this thesis ..... newline