"基于多核处理器的改进L7-Filter规则匹配算法"

The document "Improved L7-Filter's Pattern Matching Algorithm Based on Multi-Core Processors" proposes a dynamic adaptive algorithm for rule chaining on multi-core processors, taking into account the architectural characteristics of multi-core processors and the temporal locality of network data flows. By categorizing network data flows and dynamically optimizing rule chains based on the temporal locality of network data flows, the algorithm effectively reduces the number of matches for network data flows under multi-core processors, significantly improving the efficiency of rule matching. Simulation results show that the proposed algorithm offers approximately a 7% improvement in performance under the same number of network packets. As the number of network packets increases, the superiority of the algorithm becomes more pronounced. Key words: multi-core processors; network data flows; L7-Filter; temporal locality; packet classification; dynamic optimization. The research was led by authors Yu Tao and Wu Wei-Dong from the College of Computer Science and Technology at Wuhan University of Science and Technology in China.