高阶QAM信号盲均衡算法研究

发布时间：2018-08-04 16:07

【摘要】：盲均衡技术作为高阶QAM解调系统中的一项关键技术,无需周期性地发送训练序列,可仅仅依靠接收信号的统计特性来恢复原始发送信号。它在多点网络和非协作通信中具有独特的优势。但在多径信道情况下,由于高阶QAM星座点比较密集,传统Bussgang盲均衡算法有时不能有效消除码间干扰。因此,研究收敛速度更快、稳态误差更低的盲均衡算法具有重要意义。本文在分析传统盲均衡算法的基础上给出了三种改进的算法:MCME-MCMA、MMCMA-DD-VSS-LMS以及归一化解相关算法,仿真结果证明了新算法的有效性。本文主要内容如下:1.分析了无线信道特性,阐述了盲均衡算法的基本原理,详细描述了盲均衡算法性能评价准则如双模式算法的切换准则、Bussgang类盲均衡器结构。论述了分数间隔盲均衡原理和实现。并在不同场景下,对经典Bussgang算法进行了仿真实现和性能对比,且从理论上分析了各算法的优缺点。2.针对CME算法冷启动能力差和MCMA算法应用到高阶QAM信号中收敛速度较慢的缺点,提出了一种动量星座匹配算法(MCME-MCMA)。仿真结果表明,对于经过恶劣多径信道传输的64QAM和256QAM信号来说,所提算法能够快速收敛,且稳态性能比MCMA算法有5至7dB的提升。3.受MCMA-DD-LMS算法的启发,提出了一种混合盲均衡算法MMCMADD-VSS-LMS算法。在相同信道条件下,新算法在与MCMA-DD-LMS算法稳态误差相当的前提下,收敛速度提前了1000个码元。通过引入新的变步长函数增进了MCMA算法的收敛速度。应用改进算法在16QAM的盲均衡中,收敛速度相比MCMA算法提前4000个码元,稳态误差降低2dB。4.给出了一种归一化的自适应解相关算法,该算法是将解相关方法应用到盲均衡中,基于输入信号解相关前后的相关性有选择地进行解相关,用于减小均衡过程中相邻信号矢量间的相关性。结果表明:在不增加稳态误差的前提下,该算法可显著地提高均衡器收敛速度。
[Abstract]:As a key technique in high order QAM demodulation system, blind equalization can recover the original signal without sending training sequence periodically. It has unique advantages in multi-point network and non-cooperative communication. However, in multipath channels, due to the high order QAM constellation is dense, the traditional Bussgang blind equalization algorithm can not effectively eliminate the inter-symbol interference (ISI). Therefore, it is of great significance to study blind equalization algorithms with faster convergence speed and lower steady-state error. Based on the analysis of the traditional blind equalization algorithm, this paper presents three improved algorithms: MCME-MCMA-MMCMA-DD-VSS-LMS and normalized decorrelation algorithm. The simulation results show the effectiveness of the new algorithm. The main contents of this paper are as follows: 1. The characteristics of wireless channel are analyzed, the basic principle of blind equalization algorithm is expounded, and the performance evaluation criteria of blind equalization algorithm such as the switching criterion of two-mode algorithm are described in detail, and the Bussgang class blind equalizer structure is described in detail. The principle and realization of fractional interval blind equalization are discussed. In different scenarios, the simulation and performance comparison of the classical Bussgang algorithm is carried out, and the advantages and disadvantages of each algorithm are analyzed theoretically. A momentum constellation matching algorithm (MCME-MCMA) is proposed to overcome the disadvantages of poor cold start ability of CME algorithm and slow convergence rate of MCMA algorithm applied to high-order QAM signals. The simulation results show that the proposed algorithm can converge rapidly for 64QAM and 256QAM signals transmitted through a bad multipath channel, and the steady-state performance of the proposed algorithm is 5 to 3 higher than that of the MCMA algorithm. Inspired by MCMA-DD-LMS algorithm, a hybrid blind equalization algorithm, MMCMADD-VSS-LMS algorithm, is proposed. Under the same channel condition, the convergence rate of the new algorithm is 1000 symbols ahead of the steady state error of the MCMA-DD-LMS algorithm. The convergence rate of MCMA algorithm is improved by introducing a new variable step size function. In the blind equalization of 16QAM, the convergence rate of the improved algorithm is 4,000 symbols ahead of the MCMA algorithm, and the steady-state error is reduced by 2dB.4. In this paper, a normalized adaptive decorrelation algorithm is proposed, which is applied to blind equalization, and is selectively de-correlated based on the correlation before and after de-correlation of input signal. Used to reduce the correlation between adjacent signal vectors during equalization. The results show that the proposed algorithm can significantly improve the convergence rate of the equalizer without increasing the steady-state error.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.5

【相似文献】