Bio-Inspired Negative Selection Tested for Ransomware Anomaly Detection

Ransomware attacks represent a danger to the security and integrity of computer systems, resulting in considerable monetary losses and operational disruptions across various industries. Alternative methods must be developed since traditional signature-based techniques have trouble identifying new and polymorphic ransomware strains. In this paper, we examine the performance of a Negative Selection (NS) method for ransomware anomaly detection that is bio-inspired. The suggested approach is inspired by the ability of the human immune system to identify and destroy foreign substances. It uses the NS mechanism to recognize self and non-self-patterns and identify ransomware abnormalities. The system uses a population of false cells represented as binary strings and a self/non-self-conditional model to detect the existence of ransomware. We experimented with a comprehensive dataset of ransomware samples using MATLAB and WEKA to assess the algorithm’s efficacy. The findings showed that the bio-inspired NS algorithm outperformed (98.9%) compared to conventional machine learning techniques. The trials also revealed that the suggested method had a low rate of false positives and detection abilities, making it appropriate for real-time ransomware detection in various security application settings.