Research And Implementation Of Smith-Waterman Algorithm Hardware Accelerator Based On SOPC

In recent years,bioinformatics has been continuously developed and become an important discipline.Many universities and research institutes in China are vigorously developing this technology,which has made rapid development and big data.The basic precision medical treatment will become the scientific field of key development during the 13 th Five-Year Plan period in China,which makes the field of bioinformatics more concerned.Biological sequence analysis is an important basic research work in the field of modern life sciences,and the fundamental of sequence analysis is sequence alignment.At present,the genetic database of 1 million people has been established in the United States.It is expected that the genetic database of China will far exceed this magnitude,and with the rapid development of next-generation sequencing technology,the number of sequences required for comparison analysis has an explosive growth trend.However,in today's sequence alignment technology,it has been difficult to find a technique for rapidly increasing the number of matching sequences,and then the problem that the sequence matching speed cannot keep up with the speed of the sequence generation.Aiming at this problem,this paper carries out the design and implementation of the SOPC-based SmithWaterman gene sequence alignment algorithm hardware accelerator.This study uses the Smith-Waterman algorithm principle and the pulsation array idea to implement the algorithm based on the SOPC hardware system.The high-speed completion solves the above-mentioned speed mismatch problem and has very important engineering application value.In this paper,the genetic data processing flow is analyzed first,and the key steps are extracted.Through the feature analysis of the processing model,the conclusion that the sequence comparison link takes the most time is obtained.Then the hardware implementation of the Smith-Waterman algorithm is developed and the algorithm is used.The result matrix parallelism parallelism,combined with the idea of pulsating array on the circuit,proposed a parallel processing scheme suitable for SOPC hardware system;then built the hardware accelerator of Smith-Waterman algorithm,array controller with behavioral level modeling algorithm,processing unit array,sequence replenishment,backtracking and other modules,and complete the module simulation;build the software processing platform of Smith-Waterman algorithm,use C language(MFC control)to develop the sequence comparison algorithm program,use this platform to verify the hardware accelerator The correctness of the simulation;finally,the lower board verifies the actual acceleration effect of the hardware accelerator.Based on the SOPC hardware development platform independently developed by the team,the dynamic reconfigurable function is used to realize fast switching between different algorithm types.The experimental results show that the Smith-Waterman algorithm hardware system can achieve short sequence alignment of 40×50bp in this design.Comparing the running time and comparison result of the algorithm hardware accelerator with the algorithm software platform,we can find the comparison between the two.The results are consistent,and the time required to complete a sequence comparison hardware system is 1/40 of that of the software program.Finally,the hardware accelerator implements the high-speed completion of the Smith-Waterman algorithm,achieving the purpose of the design.