脉冲噪声下OFDM系统的参数估计研究

发布时间：2018-03-13 09:09

本文选题：脉冲噪声　切入点：正交频分复用　出处：《西安电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：正交频分复用(OFDM)作为一种特殊的多载波传输方案,以其高效的频谱利用率和较强的抗多径能力等特点,广泛应用于军事和民用通信中。OFDM系统的参数估计问题一直以来是通信领域的研究热点,研究该问题具有重要的理论意义和实际的应用价值。常规的OFDM信号参数估计方法将背景噪声建立为高斯模型,但是近年来的研究发现,由于实际无线通信环境中普遍存在的一些自然和人为噪声源,诸如闪电,雷击,汽车点火和外台信号等,信道噪声常常伴有显著的短时大幅度脉冲,这种噪声可用Alpha稳定分布模型精确描述。在这类非高斯噪声背景下,常规的OFDM系统的参数估计方法性能退化甚至失效。因此,本文将重点研究Alpha稳定分布噪声下OFDM系统的参数估计问题。在OFDM系统中,多普勒效应和收发两端本振频率不匹配等因素通常会引起频率偏移。由于OFDM对频率偏移非常敏感,因此,频率偏移估计是OFDM系统的关键技术之一。在Alpha稳定分布噪声中,常规频率偏移估计的方法性能退化。针对这一问题,本文提出了一种基于Myriad滤波的频率偏移改进算法。Myriad滤波是一种适用于Alpha稳定分布随机过程的非线性滤波方法,对脉冲噪声具有较好的抑制作用。因此,本文提出的算法首先采用加权Myriad滤波器对接收信号进行滤波处理,减小脉冲噪声对参数估计的影响,然后结合SC算法的频率偏移估计思想,依次估计出OFDM系统的小数倍频率偏移和整数倍频率偏移。仿真结果表明,在Alpha稳定分布噪声下,该方法具有良好的估计性能。OFDM信号的时域参数估计对通信对抗、非协作通信、无线电频谱监测等领域具有重要的意义,本文针对常规OFDM信号时域参数估计方法在Alpha稳定分布噪声环境下性能退化的问题,提出了一种基于相关熵的时域参数估计新方法。相关熵是适用于非高斯信号处理的一种广义相关函数,用于表征随机变量的局部相似性。本文提出的方法利用OFDM信号的时域结构具有局部相似性这一特点以及相关熵对脉冲噪声较好的抑制作用,完成Alpha稳定分布噪声下OFDM信号有用符号时间和符号周期这两个时域参数的估计。此外,为进一步提高强脉冲噪声下有用符号时间和符号周期的参数估计性能,本文利用累积法对该方法进行了改进。
[Abstract]:As a special multi-carrier transmission scheme, orthogonal Frequency Division Multiplexing (OFDM) is characterized by its high spectral efficiency and strong anti-multipath capability. The parameter estimation of .OFDM system is a hot topic in the field of communication, which is widely used in military and civil communications. It has important theoretical significance and practical application value to study this problem. The conventional OFDM signal parameter estimation method establishes the background noise as Gao Si model, but in recent years, it has been found that, Because of some natural and man-made noise sources, such as lightning, lightning strike, automobile ignition and outstation signals, the channel noise is often accompanied by significant short and large pulses. This kind of noise can be accurately described by Alpha stable distribution model. Under the background of this kind of non-#china_person0# noise, the performance of the conventional parameter estimation method for OFDM system degenerates or even fails. In this paper, we will focus on the parameter estimation of OFDM system under the stable distributed noise of Alpha. In OFDM system, the Doppler effect and the mismatch of the local oscillator frequency at both ends of the transmitter and receiver usually cause the frequency offset. Because the OFDM is very sensitive to the frequency offset, Therefore, frequency offset estimation is one of the key techniques in OFDM systems. In Alpha stable distributed noise, the performance of conventional frequency offset estimation methods is degraded. This paper presents an improved frequency offset algorithm based on Myriad filter. Myriad filter is a nonlinear filtering method suitable for Alpha stable distribution stochastic process. The proposed algorithm firstly uses weighted Myriad filter to filter the received signal to reduce the influence of impulse noise on parameter estimation, and then combines the frequency offset estimation idea of SC algorithm. The fractional frequency offset and integer multiple frequency offset of OFDM system are estimated in turn. The simulation results show that the proposed method has good estimation performance under stable distributed noise of Alpha signal. Radio spectrum monitoring is of great significance. In this paper, the performance degradation of conventional time-domain parameter estimation methods for OFDM signals in Alpha stable distributed noise environment is discussed. A new time-domain parameter estimation method based on correlation entropy is proposed. Correlation entropy is a generalized correlation function suitable for non-#china_person0# signal processing. This method is used to characterize the local similarity of random variables. Based on the characteristic of local similarity in time domain structure of OFDM signal and the good suppression effect of correlation entropy on impulse noise, The time domain parameters of useful symbol time and symbol period of OFDM signal under Alpha stable distributed noise are estimated. In addition, in order to further improve the performance of parameter estimation of useful symbol time and symbol period under strong impulse noise, In this paper, the cumulation method is used to improve the method.
【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】：TN911.23;TN929.53

【共引文献】