融合压缩采样与深度神经网络的直接序列扩频参数估计

Estimation of DSSS Signal Parameter Based on Compressive Sampling and Deep Neural Networks

直接序列扩频(Direct Sequence Spread Spectrum,DSSS)信号的宽频带特性所带来的高采样率增加了参数估计的实现难度。针对现有技术所面临的问题与挑战,融合压缩采样与深度神经网络(Deep Neural Network,DNN),提出了用于估计DSSS信号参数的方法。一方面,压缩采样能够利用信号的冗余性,以较低的采样率获取信号中的有效参数信息;另一方面,DNN在提取数据特征方面具有高效准确的特点。通过对压缩采样与参数估计网络的整体训练,实现两者的有效配合,进而实现以较低采样率对DSSS信号参数的准确估计。仿真实验证明了该参数估计方法在低信噪比下的估计能力相对于传统方法具有一定的提升。

The high sampling rates caused by the wideband characteristics of the direct sequence spread spectrum(DSSS)signals increase the difficulty of the parameter estimations.Regarding to the problems and challenges within the existing techniques,a method based on the combination of the compressive sampling(CS)and deep neural network(DNN)is proposed to estimate the parameters of the DSSS signals.On the one hand,the CS can obtain the information on the parameters effectively with low sampling rate,by exploiting the redundancy within signals;on the other hand,the DNN can extract the data features effectively and accurately.By the combined training of the compressive sampling and the parameter estimation network,the effective cooperation of the two parts and the accurate estimation of the DSSS estimations with low sampling rates are achieved.The simulations prove that the proposed method outperforms the conventional method in terms of estimation capability in low signal-to-noise ratio.