Abstract: For the problem of the soft spectrum signal pseudo-code sequence was difficult to estimate,a blind estimation method of soft spread spectrum signal was proposed based on improved density peak clustering.On the premise of known pseudo-code cycle and chip rate, the received data was segemented to obtain the data matrix.Then,the local density of the data points and the shortest distance from the data points with higher local density were calculated.Finally,the density peak clustering algorithm was used to automatically determine the cluster center to estimate the pseudo-code size and pseudo-code sequence of walsh soft spread siganl.The proposed algorithm has the characteristics of realizing the blind estimation of soft spread spectrum signal of walsh code under the condition of low SNR.The simulation results show that the proposed algorithm can accurately estiamte the pseudo-code sequence size and pseudo-code sequence when the SNR is slow,and the time conplexity is low.