Channel Estimation And SNR Estimation For LTE Downlink SISO System

The LTE is the next-generation mobile communication scheme, which was proposed by 3GPP. Using OFDM, MIMO as its key technologies, it goals at high mobility, high data rate and high transmission quality.In order to adapt to the complex mobile channel environment, appropriate channel estimation technology must be used. Meanwhile an accurate SNR estimation is the key to AMC, which can effectively improve the spectrum efficiency. These two technologies for LTE downlink SISO system were studied in this thesis.First, the basic knowledge of physical layer in LTE system was briefly introduced, such as frame structure, downlink slot structure, physical resource distribution, etc., followed by the design of simulation platform based on 3GPP's LTE protocols. Also, the CP selection criterion was analyzed, and urban channel model was designed as the later simulation environment in this part. Second, pilot-based channel estimation for LTE downlink was studied and simulated, including the LS, MMSE, and several interpolation algorithms. According to characteristics of pilots in LTE downlink, LS and MMSE were simplified. The third part is mainly about the LTE downlink SNR estimation technology. According to simulation, some classical algorithms were feasible in the AWGN channel but their estimation bias was great for multi-path channel. Then modified Boumard's algorithms were proposed, which were applicable to LTE downlink system, but simulation showed that there were some constraints. The improved algorithm with much better performance was given in this thesis, based on the variation regularity of channel response of sub-carriers which are close to each other in frequency domain.Finally, comprehensively considering the simulation conclusion and the complexity of implementation, a scheme for designing the LTE downlink channel estimator was presented, which was also implemented on DSP platform.