Design And Implementation Of A Hardware Search Engine Based On SRAM

Different from static random access memory(SRAM),TCAM(Ternary Content Addressable Memory)can not only access memory through address,but also input search data to get the address of corresponding data.Because of its fast search ability,TCAM is widely used in network infrastructure of data search function.However,the traditional TCAM has high power consumption and high price,so many TCAM based on SRAM have been proposed,but the resource utilization of these TCAM based on SRAM is still low.This paper proposes a resource-saving TCAM structure based on SRAM(RS-TCAM),which can effectively reduce the use of SRAM and improve the efficiency of resource utilization by dividing coding area and compressing the data of coding area.The experimental results based on Xilinx virtex-7 FPGA and real routing table show that the proposed RS-TCAM architecture uses less 10.94% SRAM at most compared with UE-TCAM.This paper also applies RS-TCAM structure to Network Intrusion Detection System(NIDS),RS-TCAM plays a "filtering" role.Through the filtering of RS-TCAM,NIDS does not need to compare the data package with all the rules in the system one by one to determine whether there is "dangerous behavior" in the data package.It only needs to compare with a few rules to determine.Therefore,the application of RS-TCAM to NIDS improves the detection efficiency of the system.In order to prove that NIDS based on RS-TCAM can improve the detection efficiency,this paper does some simulation experiments : six groups of data were intercepted for relevant experiments from DARPA2000.According to six groups of experimental data,the average filtering ratio is 93.2%.That is to say,for data packets to be detected,only the overall comparison with 6.8% of the rules was needed.Obviously,RS-TCAM plays a very good "filtering" performance in network intrusion detection system,which can greatly improve the detection efficiency of network intrusion detection system.