Research On Access Multi-specifications Signal And Detection Technology For High-speed Network Flow

With the continuous increase of network transmission rate and the increasingly rich network applications,the specifications and standards are more and more complex about the signal of the backbone network transmission.Studying the network traffic characteristics and ensuring the normal and stable operation of the backbone network are the key problems in network technology research.Network signal detection technology plays a very important role in detecting network traffic composition,understanding network operation status,maintaining network transmission security and improving network service quality.Traditional network flow detection system accesses a single specification network signal and faces the packet level for detection.It is poor adaptability,costly resources,and has been difficult to adapt to the current network signal detection technology needs.The study of multi-specification signal access network flow detection technology can solve the above problems.In this paper,the network flow defined by the five-tuple is taken as the research object,which solves the multi-specification signal access,obtains the network flow information through the protocol analysis and cache,and realizes the network flow level detection through the hardware solution.According to the result,it has a different processing strategy.This paper focuses on the whole process of high-speed network flow detection,mainly completed the following work:?1?This paper researches on the accessing technology for multi-specification signal and proposes a solution of the multi-specification access technology for Ethernet and POS signals at different rates.This solution solves the problems about the different recovery clocks and data bit widths involved in multi-specification signals access at the same time,and finally meets the actual demanding on simultaneous accessing to multi-specifications signal.?2?This paper studies on the algorithm of network flow detection,discusses the difference between time complexity and space complexity,and propose a high-speed flow detection algorithm of hardware based on XOR hash table compression and TCAM,which solves the rule table length paradox between speed and storage capacity of the TCAM flow detection algorithm process.Compare to the original algorithm,the efficiency can be improved two times at the expense 3.6379×10^-12 of the false positive rate.?3?This paper designs and implements the hardware platform of high-speed network flow detection system through the theoretical analysis and actual demand of the flow detection technology,and adopt the modular and functional design criteria,and realize multi-specifications signal accessing and the flow detection algorithm of hardware based on TCAM.?4?This passage design a targeted test program,the key functions of the hardware platform has been done a complete test,the results show that the system is running properly and can meet the requirement of the high-speed network flow detection processing.