Stream classification device and table item compression method based on folded SRAM (Static Random Access Memory)

Abstract

The invention belongs to the technical field of network communication, and discloses a flow classification device and a table item compression method based on a folded SRAM (Static Random Access Memory), which are characterized in that a rule set is preprocessed to obtain an execution action and a compression action; storing the execution action to an SRAM (Static Random Access Memory), and storing the compression action to a register vector; the analyzer receives an input data stream, extracts a key information field in header information of a data packet, completes splicing operation, and outputs the key information field in a key information vector mode; receiving a key information vector, and performing segmented storage on the key information vector; the rule matching module is based on a BV lookup mode of a folded SRAM, and an index address is corrected according to an offset value obtained by a compression action so as to complete table lookup operation; the execution action obtained by table look-up is multi-stage cyclic matching of the stream classification instruction code and the compression action until a final matching result is obtained and output. The method has the advantages of being suitable for hardware implementation, reducing cache overhead and having nothing to do with a protocol, and can be used for data flow identification and classification in a communication network.

Description

technical field [0001] The invention belongs to the technical field of network communication, and in particular relates to a flow classification device and a table entry compression method based on a folded SRAM. Background technique [0002] With the development of the network, customers need more diverse and faster and more secure services. As an important part of the switch, the data flow classification device undertakes the function of distinguishing the data flow according to the key information in the data flow, thereby providing different operations. The performance of the traffic classification module determines whether the switch can provide accurate, fast and diverse services according to customer needs. One of the kernel functions performed by OpenFlow is flow table lookup, which is similar to the classic multi-field packet classification mechanism. With the refinement of the control plane of OpenFlow, the number of matching fields in the flow table maintained b...

AI Technical Summary

Problems solved by technology

[0006] (1) The current flow classification device based on ternary content addressable memory (TCAM) has small capacity, high price and high power, and is not suitable for large-scale rule search

[0007] (2) In the current flow classification device based on the BV algorithm, the relatively large memory resource consumption seriously hinders the wide application of the algorithm

[0008] (3) The BV search device of the existing two-dimensional pipeline structure is deployed on the FPGA to implement the processing unit through logic resources. Excessive consumption of logic resources on the FPGA will affect the performance, and it is not convenient to deploy on the ASIC.

Images