高速移动场景下基于TD-LTE标准的信道估计技术

发布时间：2018-06-29 15:31

本文选题：正交频分复用 + 信道估计　；参考：《北京交通大学》2014年硕士论文

【摘要】：随着移动通信技术的不断发展,目前LTE系统已经在全球范围内开始商用。其采用了正交分频复用(OFDM)技术,可以通过快速傅里叶变换(FFT)将高速数据流调制到相互正交的N路子载波上。这样,OFDM技术将频率选择性衰落信道转化为若干个频率非选择性衰落信道,能有效地抵抗多径引起的符号间干扰(ISI)和多径衰落。在接收端,通常采用相干检测的方法来获得频域数据符号。为了能够恢复出传输信号,我们必须要正确地估计出无线信道对传输信号的影响,这就要求事先有效的估计出信道参数。因此,在移动通信系统中,信道估计是接收机设计的重要组成部分。本文首先分析了目前存在的各种信道估计方法,它们通常假设信道参数在一个OFDM符号内是恒定不变的。基于LTE系统的信道估计算法属于导频辅助估计,主要有两个步骤：首先对含有参考信号的OFDM符号进行统计估计,这里主要采用最小二乘(LS)算法和最小均方误差(MMSE)算法。降秩LMMSE算法、SVD算法、变换域降噪算法等信道估计算法大多是对上述两种基本方法优化改进而来。随后,数据载波处的信道响应可以通过内插技术得到,主要包括常值插值,一阶线性插值,二阶多项式插值,三次样条插值等等。本文还给出了这些信道估计算法基于TD-LTE标准的仿真分析。之后,本文研究了在高速移动场景中的信道估计方法。当终端高速移动时,会造成多普勒效应,并导致信道发生时间选择性衰落。因此,信道在一个OFDM符号内呈现出较为明显的变化,这时子载波间干扰(ICI)就会影响传统的信道估计方法,使估计性能明显下降。本文提出通过幂级数基的形式来建模无线信道变化的方法,将时变的信道估计转变成有限参数变量在幂级数基上的投影,这大大减少了估计信道参数的个数。理论分析及基于TD-LTE Release8的仿真结果表明,该方法能有效地跟踪和估计时变的信道参数,具有良好性能。
[Abstract]:With the development of mobile communication technology, LTE system has been used in the world. The orthogonal Frequency Division Multiplexing (OFDM) technique is used to modulate the high-speed data stream to the orthogonal N-channel subcarriers by means of fast Fourier transform (FFT). In this way, OFDM can transform frequency selective fading channels into several frequency non-selective fading channels, which can effectively resist inter-symbol interference (ISI) and multipath fading caused by multipath. At the receiving end, coherent detection is usually used to obtain frequency domain data symbols. In order to recover the transmitted signal, we must correctly estimate the influence of the wireless channel on the transmitted signal, which requires the effective estimation of the channel parameters in advance. Therefore, in mobile communication systems, channel estimation is an important part of receiver design. In this paper, we first analyze the existing channel estimation methods, which usually assume that the channel parameters are constant in an OFDM symbol. The channel estimation algorithm based on LTE system belongs to pilot aided estimation. There are two main steps: firstly, the OFDM symbol with reference signal is estimated statistically, and the least square (LS) algorithm and the minimum mean square error (MMSE) algorithm are mainly used in this algorithm. Most of the channel estimation algorithms such as reduced rank LMMSE algorithm SVD algorithm and transform domain noise reduction algorithm are improved from the above two basic methods. Subsequently, the channel response at the data carrier can be obtained by interpolation techniques, including constant interpolation, first-order linear interpolation, second-order polynomial interpolation, cubic spline interpolation and so on. This paper also presents the simulation analysis of these channel estimation algorithms based on TD-LTE standard. Then, the channel estimation method in high-speed mobile scene is studied. When the terminal moves at high speed, it causes Doppler effect and time selective fading. Therefore, the channel changes obviously in an OFDM symbol. In this case, the inter-carrier interference (ICI) will affect the traditional channel estimation methods, and the estimation performance will be significantly reduced. In this paper, a method of modeling wireless channel variation by power series basis is proposed. The time-varying channel estimation is transformed into the projection of finite parameter variables on the power series basis, which greatly reduces the number of estimated channel parameters. Theoretical analysis and simulation results based on TD-LTE release 8 show that the proposed method can effectively track and estimate time-varying channel parameters and has good performance.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.5

【参考文献】