多用户长码直扩信号的信源数估计

针对低信噪比(signal-to-noise ratio,SNR)下多用户长码直扩信号在信源数估计中容易产生虚拟用户数的问题,提出一种基于阵列接收的信源数估计方法。首先采用阵列天线对长码直扩信号进行信号接收;然后使用信息论、盖氏圆以及平滑秩的算法实现信源数的估计,最后对3种算法进行对比研究。所提的信源数估计方法相比于直接用3种算法估计信源数在信噪比上有很大的提高,而且不会产生虚拟用户数。理论分析和仿真结果表明:在信噪比较低的环境下,对周期长码和非周期长码信号的信源数估计都能达到较好的估计性能。

Estimation of the number of sources of multi-user long code DSS signals

Aiming at the problem that the number of virtual users is easy to be generated in the source number estimation under low SNR, a source number estimation method based on array reception is proposed. The method firstly uses the array antenna to receive the signal of the long code DS signal. Then, using the information theory, Gai"s circle and smooth rank algorithm to estimate the source number, the four algorithms are compared. The proposed method for estimating the number of sources has a significant improvement in the signal-to-noise ratio compared to the estimation of the number of sources directly by the four classical algorithms, and does not generate a virtual number of users. Theoretical analysis and simulation results show that both the long-term and non-periodic long-code signals can achieve good estimation performance under low signal-to-noise ratio.