Research On Pilot-based Channel Estimation In OFDM System

The sharp increase of demand for mobile communication promotes the new technology being used in wireless digital mobile communication systems. With the advantage of robustness against multi-path delay spread, high bandwidth efficiency, and simple hardware implementation, etc. Orthogonal frequency division multiplexing (OFDM) technology which is a multi-carrier modulation technology is widely used in kinds of digital communication systems. Channel estimation is the fundament of coherent detection, demodulation and equalization, and its accuracy determines the OFDM system's performance.This paper mainly studies the technology of channel estimation in OFDM system. Firstly the paper introduces the development status and application, describes the basic principles of OFDM, and introduces the achievement and key technologies of OFDM system. Then, the current characteristics of the wireless mobile channel is studied, and the multi-path rayleigh fading channel model are described in detail. In this paper, the research of pilot-based channel estimation in OFDM system is studied from two aspects.One aspect is based on channel estimation method in pilot location. At first, we discuss the pilot selection and insertion methods in channel estimation, and then analyze the channel estimation algorithms in pilot location in detail, including: least square(LS),minimum mean square error(MMSE),linear minimum mean square error(LMMSE), singular value decomposition(SVD) and discrete fourier transform(DFT)-based channel estimator. But due to the limitations of LS-DFT algorithm, this paper presents an improved algorithm based on threshold judgment, this estimation method has a more general scope of application. According to the simulation results, the proposed algorithm has better performance.The channel estimation of non-pilot can be realized by interpolation algorithms, which can get all the location of all sub-carrier channel frequency response. The algorithms of constant interpolation, linear interpolation, Gauss interpolation, Cubic interpolation and FFT-based time-domain and conversion domain are analyzed. In order to decrease the error caused by the spectral leakage of DFT calculation, two improved algorithms of this paper based on FFT interpolation is proposed, one is the improvement of frequency-domain windowing interpolation algorithm, the other is data flip interpolation algorithms based on time and frequency domain respectively. In the end, the analysis and comparison of these algorithms'performance are made by computer simulation.