非协作多用户短码直扩信号伪码估计

Pseudo-code estimation of non-cooperative multi-usershort code direct spread signal

在非协作短码直扩通信下,非整数倍码元速率采样时,多用户信号的伪码估计往往不准确。为此,提出一种结合伪码延迟自相关法和主特征独立分量分析法的伪码估计方法。首先,采用平方谱法估计多用户信号载波频率,并采用延时相乘-自相关法估计多用户信号的伪码速率;然后,运用伪码速率估计值对多用户信号进行采样率转换,并采用时域延迟自相关法估计伪码周期;最后,根据伪码周期估计值,运用伪码延迟自相关法寻找采样率转换后多用户信号中伪码信息段的起始点,提取伪码信息段,构建盲源分离模型,运用主特征独立分量分析法对建模后的伪码信息矩阵进行分离,得到多个用户的伪码。仿真实验结果表明,在非协作短码直扩通信下,该估计方法的单用户直扩信号的伪码估计误差低于滑动窗口法,多用户直扩信号的伪码估计误差低于基于ICA的多用户伪码估计方法和滑动窗口法。

In the non-cooperative short-code DSS communication,the pseudo-code estimation of multi-user signal is often inaccurate when the non-integer multiple symbol rate is sampled.Therefore,a pseudo-code estimation method combining the pseudo-code delay autocorrelation method and the principal feature independent component analysis method is proposed.Firstly,the carrier frequency of multi-user signal is estimated by the square spectrum method,and the pseudo-code rate of multi-user signal is estimated by the delay multiplication-autocorrelation method.Then,the pseudo-code rate estimation is used to convert the multi-user signal sampling rate,and the time domain delay autocorrelation method is used to estimate the pseudo-code period.Finally,according to the estimated pseudo-code cycle value,the pseudo-code delay autocorrelation method is used to find the starting point of the pseudo-code information segment in the multi-user signal after sampling rate conversion,and the pseudo-code information segment was extracted to construct the blind source separation model.The modeled pseudo-code information matrix is separated by the principal feature independent component analysis method,and the pseudo-code of multiple users is obtained.Simulation results show that under the non-cooperative short-code DSS communication,when the input SNR is-1 dB,the pseudo-code estimation error of single-user signal is reduced by 9 dB compared with the sliding window method,and the pseudo-code estimation error of multi-user signal is reduced by 6 dB compared with the multi-user pseudo-code estimation method based on ICA and the sliding window method.