The Research Of The Deep Packet Inspection Technology Based On Improved Bloom Filter

With the further enhancement of the computer network resources sharing, the issue of information security has become more prominent, which has become an important part of a nation, national defense and the national economy. To improve detection capabilities of the firewall, IDS (Intrusion Detection System), IPS (Intrusion Prevention System) and other security tools, Deep Packet Inspection technology has been in place. In the Deep Packet Inspection technology, the string matching technique has been widely used, which has become a decisive factor affecting the system performance. This dissertation, basing on this background, studied applications of an improved matching algorithm in the Deep Packet Inspection technology.This dissertation firstly analyzes the present situation, problems and trends of network security equipment, then introduces the basis of the deep packet inspection technology and Intrusion Prevention Systems, emphasizes on packet filtering technology and flow filtration technology. After that, this dissertation proposes a Deep Packet Inspection model based on string matching with a detailed description of the framework and design components of the model structure, and key issues affecting the performance of packet inspection has been studied. Based on the study and analysis of the relevant string matching algorithm, such as the classic BM algorithm and the standard BF algorithm, the thesis proposes an improved BF string matching algorithm. At last, BM, BF and the improved algorithm have been simulated in Snort system; matching speed and used memory of algorithms have been contrastively analyzed.