面向LTE系统的终端定位算法研究

发布时间：2018-08-26 21:33

【摘要】：为满足用户不断增长的定位需求,3GP PLTE制定了新的定位协议,对定位提供了全面支持,无线定位将在LTE时代的民用和军用领域内发挥重要的实用价值。但目前LTE定位技术未充分利用MIMO和OFDM技术为其提供的优势,存在以下三个方面的问题：1)传统的时延估计算法在LTE系统中性能受限；2)LTE协议中OTDOA定位算法鲁棒性较差；3)传统的NLOS误差鉴别方法在LTE系统中鉴别精度低。针对上述问题,本课题依托军队某重点项目,对LTE系统的定位技术展开研究。首先,根据LTE系统OFDM信号的时、频特性提出一种分组联合时延估计算法；然后,在估计时延的基础上,提出一种改进的Mean-OTDOA定位算法减小远近效应对定位精度的影响；最后,针对实际NLOS场景,根据LTE系统丰富的空、时、频信息提出一种NLOS鉴别的方法。具体研究工作如下：1、提出一种基于OFDM信号特性的分组联合时延估计算法。首先,根据OFDM信号的时频特性提出一种基于频域相偏的多径时延估计模型。在此基础上,结合LTE系统的实际特点提出一种分组联合时延估计算法一-GJ-TDE算法。该算法首先将OFDM系统的时域接收信号转换为频域信号,然后对频域信号抽样分成多组低维的接收数据矩阵,并利用各抽样数据矩阵组分级估计时延；最后,取各时延估计值的平均作为定位时延值。仿真结果表明：在SNR为0dB、抽样间隔为8的条件下,GJ-TDE算法的RMSE比基于时域同步的Mensing算法降低了5.5034m。2、提出一种改进的Mean-OTDOA定位算法。在得到终端与各基站的时延估计值的基础上,计算出终端与参与定位的各基站的距离测量值并对终端与各基站的距离测量值进行平均,作为OTDOA定位方法中的参考距离；然后,利用Taylor级数展开法对终端位置进行估计。仿真结果表明：在基站数目为5、测量误差的标准差为50m时,Mean-OTDOA算法的RMSE比OTDOA算法降低了5.2039m,从而提高了终端定位精度。3、提出一种基于空、时、频联合信息的NLOS鉴别方法。首先,推导出线性抽头延迟线信道模型的频域信道响应的归一化互相关函数；然后,构造出在LOS和NLOS情况下具有不同分布特点的STFSCD(空时频相关差之和)参数,并根据STFSCD的大小鉴别NLOS。仿真结果表明：在LTE系统的EPA信道模型中,利用STFSCD参数可鉴别出NLOS,且当STFSCD参数的判决门限为3时,NLOS的正确检测概率为96.88%,虚警概率为3.6%。
[Abstract]:A new positioning protocol has been developed to meet the increasing needs of users. It provides a comprehensive support for positioning. Wireless positioning will play an important role in the civil and military fields in the LTE era. However, the current LTE localization technology does not make full use of the advantages provided by MIMO and OFDM technology, and there are three problems in the following three aspects: 1) the traditional time delay estimation algorithm has poor robustness in the LTE system with limited performance / 2) OTDOA location algorithm in the LTE protocol; 3) the traditional NLOS error discriminant method has low accuracy in LTE system. In order to solve the above problems, this paper studies the positioning technology of LTE system based on a key military project. First of all, according to the time and frequency characteristics of OFDM signal in LTE system, a packet joint time delay estimation algorithm is proposed. Then, based on the estimation of time delay, an improved Mean-OTDOA localization algorithm is proposed to reduce the influence of near-far effect on location accuracy. According to the rich space, time and frequency information of LTE system, a method of NLOS discrimination is proposed for the actual NLOS scene. The main research work is as follows: 1. A packet joint time delay estimation algorithm based on the characteristics of OFDM signal is proposed. Firstly, according to the time-frequency characteristic of OFDM signal, a multipath time delay estimation model based on frequency domain phase offset is proposed. On this basis, combined with the practical characteristics of LTE system, a combined packet time delay estimation algorithm--GJ-TDE algorithm is proposed. The algorithm first converts the received signals in time domain of OFDM system into frequency-domain signals, then the signals in frequency domain are sampled into groups of low-dimensional received data matrices, and the time delay is estimated by using each set of sampling data matrices at different levels. The average of each time delay estimate is taken as the location delay value. The simulation results show that the RMSE of the GJ-TDE algorithm is 5.5034m.2less than that of the Mensing algorithm based on time-domain synchronization under the condition that the SNR is 0 dB and the sampling interval is 8. An improved Mean-OTDOA localization algorithm is proposed. Based on the time delay estimation of the terminal and each base station, the distance measurement value between the terminal and the base station participating in the positioning is calculated, and the distance measurement value between the terminal and the base station is averaged, which can be used as the reference distance in the OTDOA positioning method. The terminal position is estimated by Taylor series expansion method. The simulation results show that the RMSE of Mean-OTDOA algorithm is 5.2039mless than that of OTDOA algorithm when the number of base stations is 5 and the standard deviation of measurement error is 50m. The accuracy of terminal location is improved by 0.3.A NLOS discrimination method based on space-time and time-frequency joint information is proposed. First, the normalized cross-correlation function of the frequency domain channel response of the linear tap delay line channel model is derived, and then the STFSCD (sum of space-time-frequency correlation difference) parameters with different distribution characteristics are constructed in the case of LOS and NLOS. The NLOS. is identified according to the size of STFSCD. The simulation results show that in the EPA channel model of LTE system, the NLOS, can be identified by using the STFSCD parameter. The correct detection probability of STFSCD parameter is 96.88 and the false alarm probability is 3.6 when the decision threshold of STFSCD parameter is 3.
【学位授予单位】：解放军信息工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.5

【参考文献】