Application Research on Network Security Based on Feature Matching Technology

This paper introduces the basis deep packet inspection technology, studies data filtering technology, stream filter technology and intrusion prevention system. Then it further puts forward a string matching-based deep packet inspection model, explains framework structure of model and its various parts design in detail and analyzes key parts affecting packet detection performance. We study relevant string matching algorithm. After it analyzes classical BM algorithm and standard BF algorithm, this paper proposes an improved BF string matching algorithm. Finally, based on open source IDS system Snort, this paper performs comparison analysis on BM, BF and its improved algorithm on matching speed and memory occupation. The results prove that the improved project is superior to traditional BM algorithm on matching speed or memory occupation based on certain scale of rule set.