Sampling Frequency Offset Estimation for Pilot-Aided OFDM Systems in Mobile Environment.
Yiling Wu,Yuping Zhao,Dou Li
DOI: https://doi.org/10.1007/s11277-010-0049-x
IF: 2.017
2010-01-01
Wireless Personal Communications
Abstract:In orthogonal frequency division multiplexing (OFDM) systems, most of the conventional sampling frequency offset (SFO) estimation methods work under the assumption of time-invariant or slow time-variant channels. In mobile environment, the time-variant channel significantly degrades the accuracy of SFO estimation. To solve the problem, we first analyze the properties of time-variant channels. If terminal moves within some tens of the wavelength of radio frequency (RF) signal, channel path delay almost remains unchanged. For most practical OFDM systems, our analysis indicates that channel path delay can be regarded as unchanged during the interval of some tens of OFDM symbols in time-variant channels. Based on the analysis, we propose a novel SFO estimation method for pilot-aided OFDM systems. Different from the conventional methods, the proposed method estimates SFO by detecting the variation of the symbol timing error caused by SFO. The detection is finished by implementing correlation between the channel impulse responses (CIRs) estimated by different OFDM symbols. Performance of the proposed method is simulated and compared with that of two conventional post-FFT methods. Numerical results show that, the proposed SFO estimation method performs better than the conventional methods not only in time-variant channels, but also at low SNRs and large residual carrier frequency offsets (CFOs).