Descriptive analysis of Hash Table based Intrusion Detection Systems

Security and the confidentiality during the data transfer are the important metric in the network design. A group of sequential actions to assure the data confidentiality refers the intrusion. Intrusion in network gathers the information related to unauthorized access, and the exploitation of several vulnerabilities raised by attacks. This paper presents the detailed survey of strategies involved in the implementation of Intrusion Detection Systems (IDS) in the network. The survey categorized into five phases namely, IDS, data mining based IDS, multi-agent based IDS, Distributed Hash Table (DHT), and Internet Protocol (IP) based hash table. First phase discusses the structure of IDS with machine learning techniques such as Bayesian classifier, knowledge based, etc. Second, a data mining based IDS conveys how the reliability and security of IDS are improved compared to previous IDS. In the third phase, multi agent based IDS presents the status of coordination issues, false alarm rates and detection rates on application of multiple agents. Finally, a hash table mechanisms (Distributed Hash Table (DHT) & Internet Protocol (IP) based hash table) into the network to improve the matching efficiencies and computational speed. This survey conveys the difficulties in the traditional methods, namely, storage overhead, less matching efficiency, and adaptive nature (dynamically updating of hash tables) and false positive rates. The prediction of attackers or mis-behaving requests and the construction of adaptive reputation constitutes the main problems in IDS that lead to less efficiency. The observation from the survey lead to the stone of extension of Distributed Hash Table (DHT) with fuzzy based rules in order to overcome the difficulties in traditional research works.