Research On UE Location Algorithm In LTE Systems

To meet the growing demand for positioning,3GPP LTE makes the new positioning protocol, and supports localization fully. Wireless positioning will play an important and practical role in both the civilian and the military fields in the LTE era. But the LTE positioning technology does not take full advantage that MIMO and OFDM technologies provide, and still faces the following problems:1) the traditional delay estimation algorithm is limited in the LTE system; 2) the OTDOA algorithm is poor robustness in LTE protocol; 3) the traditional NLOS identification precise is low in the LTE system. Aimed at the above problems, the paper launches a special study on the positioning technology in the LTE systems, which is supported by a military project.Firstly, a GJ-TDE (Grouped Joint Time Delay Estimation) algorithm is proposed based on the time-frequency characters of the OFDM signals in the LTE systems. Secondly, a Mean-OTDOA positioning algorithm is proposed to reduce the near-far effects to increase the positioning accuracy; lastly, in the actual NLOS scenario, a NLOS identification method is proposed based on the space-time-frequency correlation in LTE time-varying channels. Specific research works are as follows:A GJ-TDE (Grouped Joint Time Delay Estimation) algorithm is proposed based on the time-frequency characters of the OFDM signals. An FDP-MDE (Multipath Delay Estimation Based on Frequency Domain Phase-offset) model is proposed in this paper. And then a GJ-TDE algorithm is proposed according to the features of the LTE systems. Firstly, the time domain signals are converted into the frequency domain signals. And then they are sampled into the low-dimensional received signal matrix groups, which are used to estimate the time delay and the mean of all the delay estimations is used to position finally. Simulation results show that:the RMSE of the GJ-TDE algorithm is 5.5034m smaller than that of the synchronization Mensing algorithm when the SNR is OdB and the sampling interval is 8.An improved Mean-OTDOA location algorithm is proposed. Firstly, we estimate the distance between the terminal and base stations based on the estimated time delay, then we average the measured distance as a reference distance of the OTDOA positioning method, and finally we use the Taylor series expansion method to estimate the location of the terminal, and analyze the algorithm performance. Simulation results show that:this algorithm can improve the positioning accuracy of the terminal. When the number of the base station is 5 and the standard deviation of the measurement error is 50m, the RMSE of the proposed algorithm is 5.2039m smaller than that of OTDOA algorithm.A new NLOS identification method is proposed based on space-time-frequency correlation. Firstly, we derivate the normalized cross-correlation function of the frequency-domain channel response in a tapped-delay-line model, and then construct the parameter STFSCD which is different in LOS and NLOS conditions to identify NLOS according to the value of STFSCD. Simulation results show that:NLOS can be identified by the STFSCD in the EPA channel model of the LTE systems and when the threshold of the STFSCD parameter is 3, the detection rate of the NLOS identification method is 96.88%, and the false alarm rate is 3.6%.