ID-SOMGA: A Self Organising Migrating Genetic Algorithm-Based Solution for Intrusion Detection

Abstract

The study examined the detection of attacks against computer networks, which is becoming a harder problem to solve in the field of Network security. A problem with current intrusion detection systems is that they have many false positive and false negative events. Most of the existing Intrusion detection systems implemented depend on rule-based expert systems where new attacks are not detectable. In this study, optimization algorithms were added to intrusion detection system to make them more efficient. Self Organizing Migrating Genetic Algorithm (SOMGA) was integrated into intrusion detection system to obtain a more efficient intrusion detection system called ID-SOMGA. This study provides an equally efficient method to implement an intrusion detection system that returns very low false positives. Due to the complexities involved in security issues, and the implementation of the work, selected values of the network log was used to implement the system in order to reduce some of these complexities. The Self Organizing Migrating Genetic Algorithm -Intrusion Detection System was tested and values of the result were compared with that of an IDS with Genetic Algorithm Intrusion Detection System. In terms of detection rates, ID-SOMGA was found to be slower than an IDS with GA, the false positives in ID-SOMGA was lower than what obtains with genetic algorithm. Both schemes were able to identify new patterns almost in the same way. The ID-SOMGA system that was developed improved the security of systems in networked settings allowing for confidentiality, integrity and availability of system resources.

Factors Influencing Knowledge and Persuasion of Financial Regulators in the XBRL Adoption Process: The Technological Perspective

“…Relative advantage is a form of drive that encourages technology adoption (Iacovou, Benbasat & Dexter, 1995). The literature on information systems shows that the more an organisation presumes the probable advantage of a technology, the more likely the organisation would initiate adoption (Folorunso, Vincent, Adekoya & Ogunde, 2010;Al-Jabri & Sohail, 2012).…”

Improving Bee Algorithm Based Feature Selection in Intrusion Detection System Using Membrane Computing

“…In all, feature selections are often grouped into two, namely; filter and wrapper approaches [5,6,12]. By adopting a filter is a classifier agnostic pre-selection methods which are independent of the applied machine learning algorithm.…”

“…Furthermore, Intrusion Detection System (IDS) according to [5,6,12], is viewed as a system which constantly and dynamically monitors a system with the primary aim of identifying and reporting unusual activities or actions which are symptomatic of an attack. If these supposedly attacks are not checkmated, the aftermath on a system would be: (i) compromise of its integrity, (ii) denial of its availability and (iii) inefficiency.…”

Advocating the use of fuzzy reasoning spiking neural P system in intrusion detection

“…These fields were selected based on the heuristic analysis of the data whereby they were found to be unique to the attack. So, with the assumption that all rules have full confidence factor, the formulated rules include: (2,14), (2,16), (3,15), (4,17), (5,13), (5,14), (6,15), (7,16), (7,17), (13,8), (14,9), (15,10), (16,11), (17,12)} 5) in = {σ 1 ,σ 2 σ 3 ,σ 4 σ 5 , σ 6 ,σ 7 } out = {σ 8 ,σ 9 σ 10 ,σ 11 σ 12 } Fig. 3.…”