高速卫星两级并行CMA-LMS自适应均衡

Two Stage Parallel CMA-LMS Adaptive Equalization in High Speed Satellites

为了克服高速宽带卫星传输非线性群时延信道等导致的信号传输码间干扰问题,提出了一种收敛快且稳态误差小的部分并行两级双模结构常模-最小均方(CMA-LMS)级联自适应均衡算法。该算法利用CMA及LMS算法,建立了两极双模盲-非盲混合均衡,并在频域采用了快速傅里叶变换(FFT)的快速部分并行滤波结构,以少量训练字占用微小带宽的代价,实现了高速并行高速宽带卫星传输非线性群时延信道均衡,提高了信号频带利用率,具有较好应用价值。仿真表明了该算法具有接近理想均衡滤波的较好性能。此外,还设计了采用部分并行结构的算法框架,并分析了其复杂度,使其可在资源有限约束下,实现高速可靠卫星非线性信道的频域均衡。

In order to overcome the problem of inter-symbol interference in signal transmission caused by the effect of non-linear group delay in high speed broadband satellite transmissions, a partially parallel two-stage two-mode CMA-LMS adaptive equalization with fast convergence speed and small stable error is proposed in this paper. By the CMA and the LMS algorithm, a two-stage hybrid two mode blind-or-not frequency domain equalization is set up. In addition, a FFT based fast partially parallel filter structure is adopted to implement the nonlinear group delay equalization in high speed broadband satellite transmission channel at the cost of trivial bandwidth occupation of a small amount training words. So it improves the efficiency of the signal spectrum, which enable it rather good application value. Simulation results indicate that the proposed equalization algorithm approaching ideal equalization. Furthermore, a algorithm framework of a partially possesses good performance of parallel structure is designed and analyzed with the complexity. Therefore, it can realize frequency domain equalization of the high speed and reliable non-linear satellite channel on condition that limited resources are available for the satellite communication implementation.