Evaluation of Intrusion Detection Systems in Cyber Security using Fuzzy OffLogic and MCDM Approach

Abstract

Modern cybersecurity infrastructures rely heavily on Intrusion Detection Systems (IDS) to detect 
and prevent malicious activities and unauthorized access. Given the growing complexity of 
network topologies and the rising frequency of cyber threats, evaluating IDS solutions requires a 
systematic and unbiased approach. In this study, thirteen widely used IDS models are assessed 
using a multi-criteria evaluation framework across four key dimensions: detection accuracy, 
resource efficiency, scalability, and false positive rate. The goal is to support informed, data
driven decision-making for stakeholders such as policymakers, IT administrators, and security 
analysts when selecting an appropriate IDS. The VIKOR method is employed to rank the IDS 
alternatives based on the assigned weights, while Fuzzy OffLogic is applied to integrate expert 
assessments expressed as intervals. The results reveal that modern AI-based IDS models 
demonstrate strong performance in scalability and resource utilization, and they outperform 
traditional systems in adaptability and detection accuracy. 

DOI: 10.5281/zenodo.15265874