5G系统中基于解调参考信号的信道估计方法

5G系统使用解调参考信号(Demodulation Reference Signal,DMRS)用于下行信道估计,工程中通常采用插值法得到数据位置信道响应,考虑实际信道中噪声的影响,频域采用线性最小均方误差(Linear Minimum Mean Squared Error,LMMSE)插值算法。针对LMMSE需要获取信道先验统计特性以及存在矩阵求逆运算量大问题,利用信号时域内能量集中特点估计出信道均方根时延、信噪比和信道自相关矩阵,对于DMRS信号在频域上间隔较大、DMRS信号所在子载波相关性不强的问题,通过引入导频加密方法提升DMRS信号所在子载波间的相关性以提升信道估计性能,并通过滑动窗口插值方法进一步降低LMMSE算法求逆运算复杂度。仿真结果表明,所提算法均方误差和误码率总体均优于线性插值和基于矩阵奇异值分解的LMMSE算法,并与传统LMMSE算法相比性能极为接近,而且复杂度降低了99.85%,适合实际工程应用。

A Channel Estimation Method Based on Demodulation Reference Signal in 5G System

5 G systems use a demodulation reference signal(DMRS)for downlink channel estimation.In engineering,interpolation is usually used to obtain the data position channel response.Considering the effect of noise in the actual channel,the frequency domain uses linear minimum mean square error linear minimum mean squared error(LMMSE)interpolation algorithm.For the problem that LMMSE needs to obtain the channel prior statistical characteristics and the matrix inversion calculations is large,the energy concentration characteristics in the time domain of the signal is used to estimate the root mean square delay,signalto-noise ratio and channel.The autocorrelation matrix is used to solve the problem that the DMRS signals are widely spaced in the frequency domain,and the subcarriers where the DMRS signal is located are not highly correlated.The pilot encryption method is used to improve the correlation between the subcarriers where the DMRS signal is located.The sliding window interpolation method is used to further reduce the complexity of the LMMSE algorithm inversion operation.Simulation results show that the mean squared error and bit error rate of the proposed algorithm are generally better than that of linear interpolation and LMMSE algorithm based on singular value decomposition.Compared with that of the traditional LMMSE algorithm,the performance of the proposed algorithm is very close,and the complexity is reduced by 99.85%,which is suitable for practical engineering applications.