Acceleration method, device and storage medium for deep message recognition

Abstract

The invention discloses an acceleration method, device and storage medium for deep message identification. The embodiment of the present invention uses the programmable hardware accelerator card to complete the deep message identification (DPI) function, including: first, using the network port of the programmable hardware accelerator card to receive the network message; next, using the chip of the programmable hardware accelerator card to collect Relevant information of network packets, and perform in-depth message identification on these information, such as message statistics, regular expression matching, etc.; after that, the results of in-depth message identification are connected to the host computer through a programmable hardware accelerator card sent back to the host on the PCI bus. Since the above methods are all completed on the programmable hardware accelerator card, the processing speed will be very fast, which can improve the performance of deep message recognition and the accuracy of statistical results. And because the resources of the host CPU are no longer occupied, the host CPU can also save more resources to handle more tasks, thereby greatly improving the processing capacity of the host.

Description

technical field [0001] The invention relates to the field of network communication, in particular to an acceleration method, device and storage medium for deep message identification. Background technique [0002] Deep Packet Inspection (DPI) is a packet-based in-depth inspection technology that performs in-depth inspections on different network application layer loads (such as HTTP, DNS, etc.), and determines its legality. [0003] Among them, the payload detection of the message requires a series of data analysis, processing, detection, screening and statistics, and so on. At present, most of the above operations are processed by using the CPU of the server through software methods, which puts forward higher requirements on the performance of the CPU of the server. [0004] In the 5G application scenario, the data flow is very large and the processing time limit is very high. If the above operations are still completed by software methods, it will be difficult to meet th...

AI Technical Summary

Problems solved by technology

[0004] In the 5G application scenario, the data flow is very large and the processing time limit is very high. If the above operations are still completed by software, it will be difficult to meet the application requirements.

Images