Asynchronous Parallel Finite Automaton: A New Mechanism for Deep Packet Inspection in Cloud Computing

Abstract

Security is quite an important issue in cloud computing. The general security mechanisms applied in the cloud are always passive defense methods such as encryption. Besides these, it's necessary to utilize real-time active monitoring, detection and defense technologies. According to the published researches, Deep Packets Inspection (DPI) is the most effective technology to realize active inspection and defense. However, most of the works on DPI focus on its performance in general application scenarios and make improvement for space reduction, which could not meet the demands of high speed and stability in the cloud. Therefore it is meaningful to improve the common mechanisms of DPI, making it more suitable for cloud computing. In this paper, an asynchronous parallel finite automaton (FA) is proposed. The applying of asynchronous parallelization and heuristic forecast mechanism decreases the time consumed in matching significantly, while still reduces the memory required. Moreover, it is immune to overlapping problem, also enhancing the stability. The final evaluation results show that asynchronous parallel FA has higher stability, better performance on both time and memory, and is more suitable for cloud computing.