【作者】 张宇；

【导师】 张乃通；

【作者基本信息】 哈尔滨工业大学 ， 信息与通信工程， 2010， 博士

【摘要】 最大限度的提高频谱利用效率是下一代无线移动通信的目标之一。在带宽和功率受限的无线通信中,利用多输入多输出(Multi-Input Multi-Output, MIMO)技术(多天线技术)能获得更高的频谱效率,同时正交频分复用(Orthogonal Frequency Division Multiplexing, OFDM)技术具有抵抗多径衰落和降低宽带接收机复杂度的特点,因此,MIMO技术和OFDM技术的结合(即MIMO-OFDM技术)更能够适应高信道容量、高比特信息速率等宽带多媒体应用的需求,被认为是第四代移动通信系统中的主流技术,也是当前无线移动通信领域的研究热点。同步问题是通信信号处理中首先需要解决的问题,直接关系到后续基带信号处理算法的性能,MIMO-OFDM系统对同步误差非常敏感,很小的载波频偏都将导致系统性能的急剧下降,因此准确的符号定时和精确的载波频偏估计是MIMO-OFDM系统正常工作的前提。同时,信道估计技术是实现相干检测和空时解码,使得系统在低信噪比的环境下依然具有低误码率的必要条件。基于此,本文针对MIMO-OFDM系统的信道估计和同步问题进行了相关研究,主要内容及取得的成果如下:第一,对快时变环境下MIMO-OFDM系统信道估计算法的研究。给出了MIMO-OFDM系统在时变条件下的信号模型,将基于插值的SISO-OFDM系统的时变信道估计算法推广得到了适合MIMO-OFDM系统的时变信道估计算法,并且提出了一种基于低通滤波的适用于快速时变环境下MIMO-OFDM系统的信道估计方法。该算法的基本思想是通过信道频率响应相对于导频子载波变化缓慢,而相对于噪声成分变化很快的特点,实现信道频率响应和噪声成分有效的分离,利用时域低通滤波的方法对噪声进行了有效的抑制,从而提高了信道参数估计的精确性。第二,对基于空时编码的MIMO-OFDM系统信道估计算法的研究。针对基于空时编码的MIMO-OFDM系统使用最小均方(LMS)或最小二乘(RLS)自适应信道估计算法性能差的缺点,提出了一种适用于空时分组编码MIMO-OFDM系统的两步信道估计方法。第1步运用最小方均误差估计算法对导频信道进行初始信道参数估计;第2步利用第1步得到的初始信道参数,运用提出的自适应估计算法来获得更精确的后续信道参数,从而进一步提高了信道估计的精度。理论分析和仿真表明在不同的多普勒频移的情况下,与完全采用MMSE信道估计及第二步运用LMS或RLS自适应信道估计算法相比,所提出的两步信道估计方法的MSE性能、SER性能都有明显的提高。第三,对基于导频的MIMO-OFDM系统同步算法的研究。从MIMO-OFDM信号模型出发,分析了同步误差对系统性能的影响。在此基础之上,为了克服基于传统训练序列定时同步的平顶效应及多尖现象和提高系统频谱的利用率,构造了MIMO-OFDM系统中一种倒序结构的同步训练序列并且提出了对应的定时和频偏估计方法。与传统算法相比,该算法能够进一步提高定时估计的精准度。此外,该算法只需要一个训练序列,就可以有效地对整数倍与小数倍频偏进行估计,与需要2个训练序列的传统算法相比,减少了训练序列的数量,有利于提高系统频谱的利用率。仿真结果表明,在AWGN和改进的Jake模型信道环境下,所提出的方法与Minn方案和Park方案相比在MSE性能、BER性能上都有明显的提高。第四,对基于IEEE802.11a前导结构的MIMO-OFDM系统同步算法的研究。简单介绍了IEEE802.11a的物理层标准,重点讨论了IEEE802.11a标准的符号定时算法。在此基础上,针对基于传统IEEE802.11a前导结构的符号定时同步精度不高的缺点,提出了一种适合基于IEEE802.11a的MIMO-OFDM系统的符号定时同步方法。该方法将传统正交前导结构中最后一个短训练序列的前半个周期进行了相位反转,使得基于改进的前导结构的接收信号与自身周期延时信号的互相关结果中产生一个明显的下陷,因此这可以在多径环境下更准确的获得短训练序列与长训练序列的分界点,从而有效的提高多径环境下MIMO-OFDM系统符号同步的精度。仿真结果表明,提出的符号定时方法能够准确的完成定时同步,比传统的符号定时方法的精度要高得多。更多还原

【Abstract】 The target of next generation wireless mobile communication is to obtain higher efficiency of spectrum. In the wireless communication with the limitation of bandwidth and power, the implementation of MIMO technique can achieve high efficiency of spectrum. At the same time OFDM technology has the property of high tolerance to multi-path fading and low complexity to receiver design, therefore, the combination of these two technologies, that is MIMO-OFDM, can further be efficient for wideband multimedia transmission because of its high ability of anti-fading, high channel capacity and high bit-rate data, and widely be regarded as one of the most promised techniques for 4G wireless communication, which is also to be the focus research field in the future wireless mobile communication.Synchronization is the primary procedure for signal processing, which influences the performance of following baseband signal processing. MIMO-OFDM system is sensitive to synchronization errors. Very small carrier frequency offset (CFO) may degrade system performance severely. Correct symbol timing and precise CFO estimation are necessary for MIMO-OFDM systems.At the same time, channel estimation plays a very important role in MIMO-OFDM system and that is required for coherent detection and space-time decoding in MIMO-OFDM systems. Channel estimation and system synchronization of wireless MIMO-OFDM system are discussed in this dissertation based on the above consideration. The main achievements are listed as follows:1. Based on the time-varying channels, an interpolation-based MIMO-OFDM channel estimation algorithm is discussed. The system model of MIMO-OFDM over time-varying channels is acquired, and a novel channel estimation method which adapt to time-varying channels is proposed. Because channel parameters change rather slowly with respect to subcarriers, and the intercarrier interference (ICI) and additive white Gaussian noise (AWGN) in the received signal appears as fast-changing function of subcarriers, therefore, the ICI and AWGN in the pilot subcarriers are reduced by lowpass filtering which employed by the proposed method, the performance of the system is effectively improved.2. In order to improve the performance of space-time block coded MIMO-OFDM channel estimation, a two-step channel estimation method is proposed in this paper. The Minimum Mean Square Error (MMSE) channel estimation algorithm is used to estimate initial channel parameter values in the first step and these values are used to obtain more accurate values by the proposed method in the second step, so the performance of channel estimation is improved. Simulation results show that the performance of the two-step channel estimation method greatly outperforms compared to the LMS channel estimation algorithm and the RLS channel estimation, while largely decreases the computational complexity and consumption.3. Starting from the time-domain signal model of MIMO-OFDM, this paper analyzes the influences of all kinds of errors on the received signal. Then, in order to improve the accuracy of the timing synchronization which obtained by using the conventional method and the utilization of the frequency spectrum, an efficient timing and frequency offset estimation method for MIMO-OFDM systems is presented. This algorithm constructs a special training sequence to gain a better symbol timing performance, and meanwhile requiring only one training symbol, this algorithm can efficiently estimate integer frequency offset and fraction frequency offset. Comparing with conventional algorithms, the proposed method reduces the number of the training symbols, and improves the utilization of the frequency spectrum. Simulation results show that with channel AWGN and Jake, the scheme gives very accurate estimates of symbol timing and carrier frequency offset and provides a very wide range of acquisition for the carrier frequency offset, outperforming other available methods.4. IEEE.802.11a Physical Layer Standard is introduced, and the symbol timing algorithm which is suit for IEEE802.11a is discussed. Then, in order to improve the accuracy of the symbol timing synchronization which obtained by using the conventional method, a modified preamble structure for MIMO-OFDM symbol timing synchronization is proposed. The proposed method creates a sharp fall in the delayed correlation of the samples at the last short preamble period by phase reversing the first half of the last short preamble samples and hence helps more reliable estimation of the end of the short preamble or the start of the long preamble boundary especially under high time dispersive channel. Analytical and simulation results show that the proposed timing algorithm can fulfill timing synchronization and has a higher precision than the traditional method.更多还原