【作者】 李军；

【导师】 廖桂生；

【作者基本信息】 西安电子科技大学 ， 信号与信息处理， 2005， 博士

【摘要】 未来的宽带无线通信系统的目标是实现无所不在的、高速、可靠的移动多媒体传输。正交频分复用(OFDM)技术、单载波频域均衡技术(SC-FDE)和多天线技术都是有助于达到这一目标的有效技术。在多径衰落信道下的同步和信道估计是宽带无线通信系统中需要解决的重要问题。随着集成电路技术的发展,芯片的处理能力飞速增长,一些较为复杂的算法也可以实现,这使得人们有能力进一步追求高的性能。未来无线通信系统希望在不同的无线环境中参数化的解能达到或接近优化性能。从系统设计最优化的角度考虑,对各相关参数应该联合估计。同时,在不同情况下参数估计所能达到的性能极限也是一个十分重要的问题。本论文主要针对宽带无线通信系统接收机,设计和分析定时同步、载波频偏和信道参数单独或联合估计算法,并研究了一些情况下估计这些参数所能达到的极限性能。概括起来,本文的主要内容如下: 1.针对类ESPRIT盲频偏估计方法在小样本下性能急剧恶化,提出了一种基于传播算子的OFDM载波频偏盲估计的闭式解。该算法用构造的数据矩阵计算出传播算子,然后借助于传播算子推导出频偏估计的表达式。为了在小样本时仍能有效估计,用对角加载技术对传播算子的计算公式作了修正。仿真表明,无论在高斯白噪声(AWGN)信道还是多径衰落信道中,该算法在即使1、2个码元样本时,都有较好的估计性能,克服了类ESPRIT算法需要较多样本做统计的不足。同时,与类ESPRIT相同,该算法的估计范围是整个OFDM有用信号带宽,因此具有频偏估计范围大的优点。 2.提出了多径衰落信道下,基于最大似然(ML)准则的OFDM频偏、定时和信道参数联合估计代价函数。通过对代价的进一步推导求解,得出了不同情况下的三个估计算法,这些算法都是基于导频的。首先通过推导,得到了一种整数频偏(IFO)、定时和信道联合估计算法,该算法用FFT和一维搜索同时获得IFO和定时参数,然后进一步获得信道参数,在信道阶数未知的情况下,我们对算法做了调整,用循环前缀(CP)长度代替信道阶数,仿真显示,这种代替对最终性能影响不大;其次,在无载波频偏情况下,对以上算法简化,得到了一种多径衰落信道下定时估计算法,在信道阶数未知情况下,我们用在码元前端补零的方法有效消除了用CP长度代替信道阶数产生的定时测度平台;最后,我们通过把导频码元设计成相同的两部分,提出了一种FFO和定更多还原

【Abstract】 The future broadband wireless communication systems are expected to provide ubiquitous, high-reliability, and high-data-rate mobile multimedia transmission. Orthogonal frequency division multiplexing (OFDM), single carrier frequency domain equalization (SC-FDE) and multiple antennas technique are efficient techniques to achieve this goal. The estimation of the synchronization and channel parameters with multipath fading is a crucial task in broadband wireless comunication systems. With the development of the integrate circuit technique, the ability of processing of the chips are improved rapidly and some complicate algorithms can be implemented online. As a result, more high-performance algorithms with appropriate complexity can be pursued. In the future broadband wireless communication systems, it is hope that the parametrizaed solutions yield optimal or close to optimal performance in different radio conditions. From the point of view of the optimization of entire systems, the correlated parameters should be estimated jointly. Furthermore, the limiting performance of estimation in various environments is also the important issue which need be addressed in broadband wireless communications. This dissertation is devoted to the design and analysis of timing synchronization, carrier frequency offset (CFO) and channel parameter estimation algorithms in broadband wireless communications. The main contributions of the dissertation are as follows:A propagator based closed-form blind CFO estimation algorithm for OFDM systems is proposed. This method utilizes the propagator obtained from data matrix and the diagonal loading technique to make it possible to have better performance even if only using one or two OFDM blocks. Furthermore, the range of the CFO estimation which can be handled is overall transmission spectral. Simulation results show that the performance of proposed algorithm is better than that of ESPRIT-like algorithm in the case of small samples.We present a maximum likelihood (ML) based cost function which can estimate timing, CFO and channel jointly for OFDM systems and derive three estimation algorithms in different cases from it. The algorithms are based on pilot subcarriers. First, joint IFO, timing and channel estimation is considered. Both the IFO and the timing offset can be obtained by using the fast fourier transform (FFT) and one-dimension (1-D) search and the channel estimation is followed. In the case of unknown channel order (UCO), we replace the channel order by the length of cyclicprefix (CP). The simulation results show that the replacement of channel order by the length of CP leads to the negligible loss in terms of the word error rate (WER) of systems. Secondly, we simplified the algorithm when there is no CFO. This leads to a novel timing estimation algorithm in multipath fading. In the case of UCO, we employ the zero-padding (ZP) in the front of OFDM symbol to eliminate the wide plateau caused by replacement of channel order. Finally, we design the pilot as two identical halves in time domain and abtain a joint FFO and timing estimation algorithm. We find that the essential of FFO estimation of Schmidl & Cox algorithm (SCA) is ML estimation. Computer simulations show that the estimation performance of the proposed algorithms outperforms some others.A maximum likelihood (ML) based algorithm is proposed to jointly estimate the frequency-selective channels and the CFO in MIMO-OFDM by using pilot subcarriers. The proposed algorithm is capable of dealing with the CFO range overall transmission spectral. Furthermore, we present a scheme for tracking the CFO and channel parameters in time-vary channel by exploiting the finite alphabet (FA) property of the transmitted symbols. The cases with timing error and unknown channel order are also discussed. The Cramer-Rao Lower Bound (CRLB) for the problem is developed to evaluate the performance of the algorithm. Computer simulations show that the proposed algorithm can exploit the gain from multi-antenna to improve effectively the estimation performance and achieve the CRLB in high signal-noise ratio (SNR). The proposed tracking method has also high tracking performance in typical urban channel model.Subspace-based channel estimation methods for ZP-SC-FDE systems by exploiting the zero-padding are developed. The methods can estimate frequency-selective channel blindly by exploiting the zero-padding. Sufficient condition for identifiability is presented. The estimation algorithms in frequency domain and time domain are presented respectively and the equivalence of two algorithms is also proved. The performance of the method is better than that of conventional oversampling methods, and it is robust to over-estimated channel order.We develop the CRLB of CFO in smart antenna systems and analyze the contributions of the numbers of both the array elements and pilot symbols for pilot-assisted CFO estimation quantitatively. A new algorithm for pilot-assisted CFO estimation in the presence of co-channel interference (CCI) is presented with the help of adaptive array weighted structure. However, neither adaptive weight nor array manifold was needed in the final algorithm. Furthermore, computer simulations showthat the proposed algorithm can fully exploit the gains from elements and pilots and reject the effect of CCI effectively更多还原