Abstract: Duplicate data of cloud storage system is taken as one of a large amount of redundant data,and the effective and timely remove can guarantee the stability and operation of cloud storage system.Because of the interference of data,the SNR is low,the traditional method has false peaks in the fractional Fourier domain,and it cannot effectively detect and remove the duplicate data.An improved duplicate data remove algorithm of cloud storage system was proposed based on fractional Fourier transform cumulant detection.Firstly,the delete system architecture for cloud storage system was taken,the fitness function of data storage point was defined,and system subset random probability distribution function of the cloud storage node was gotten.The constraint function was used for blocking the calibration data of storage nodes,the detection of duplicate data removing processing was taken,and the fractional Fourier transform was used to preprocess the residual signal filtering in cloud storage system.The 4 order cumulanted slice post operator was used to divide each file into blocks.To delete each file block,duplicated data detection post filtering was obtained,and data storage resource detection and deletion were realized.Simulation results show that this algorithm can improve the utilization efficiency of cluster cloud storage system resource,and duplicate data can be accurately removed with higher rate.It can effectively avoid the error removing caused by interference and leakage removing,and it has superior performance.

Key words: Fractional Fourier transform ,Cloud storage,Duplicate data