Research On The Channel Estimation Technique Based On LTE Uplink

LTE,"3G Long Term Evolution", as a project to fill the huge technology gap between the3rd and4th generation mobile communication system, has become a new development of the mobile communication system. The channel estimation is intended to make some kind of estimate error minimized while minimizing the complexity of the algorithm, and also can be realized. For LTE system, channel estimation is one of the core technologies that will directly determine the style and performance of the entire system.This paper is focused on depth study of the time-domain technology. First of all, the generation process and the link structure of the LTE uplink shared channel (PUSCH) is briefly discussed.Next, the classic pilot assisted channel estimation method, least squares (LS) method, the minimum mean square error (MMSE), and the common frequency domain or transform domain noise reduction method is analyzed.Secondly, the paper focuses on the time domain algorithm in view of the LS algorithm ignoring the noise and interference which lead to the problem of imprecise estimate. The time domain correlation algorithm by selecting the appropriate window function and threshold in processing method to suppress noise obtains the accurate pilot channel estimation value, and then the channel estimation of data location by interpolation is obtained.Again, combined with the time-domain correlation algorithm and the noise reduction methods in frequency domain or transform domain, the paper proposes two improved time-domain algorithms, that are the time-domain correlation algorithm of hanning window in frequency domain and the time-domain correlation algorithm of simplified DCT (Discrete Cosine Transform) domain.Finally, according to the specification of the LTE uplink system, using MATLAB to structure the PUSCH link, this paper looks for the optimal threshold in the algorithm of post-processing part, a key part of the improved algorithm. The conclusion is that when the first algorithm is selected the threshold will select5, but when the second algorithm is selected the threshold will select4or2.5. Then two improved algorithms are simulated and analyzed under the optimal post-processing threshold conditions. The results show that such improved algorithms can improve the estimation performance of the time domain correlation method and verify the feasibility of the modified part.