Sequence Alignment Algorithms for Intrusion Detection in the Internet of Things

Abstract

The paper reviews the intrusion detection approach based on bioinformatics algorithms for alignment and comparing of the nucleotide sequences. Sequence alignment is a nature-close computational procedure for matching the coded strings by searching for the regions of individual characteristics that are located in the same order. A calculated rank of similarity is used instead of equity checking to estimate the distance between a sequence of the monitored operational acts and a generalized intrusion pattern. Multiple alignment schema is more effective and accurate than the Smith–Waterman local alignment due to ability to find few blocks of similarity. In comparison with a traditional signature-based IDS, it is found that the nature-inspired approach provides the better work characteristics. The experimental study have shown that new approach demonstrates high, 99 percent, level of accuracy.