Orthogonal Frequency Division Multiplexing (OFDM) has been adopted in many military communications systems in recent years. Blind estimation of OFDM signals is therefore becoming imperative in surveillance and reconnaissance applications. A captured signal is oversampled for complete reception and correct estimations. Cyclostationarity of the signal is therefore introduced after oversampling. In this paper we propose a cyclostationarity-based two-step progressive algorithm to precisely estimate the sampling rate of an oversampled OFDM signal with low computational complexity. In the first step, a frequency segment of the cyclo-spectrum which contains the interested cyclic frequency is roughly determined by a coarse estimation using small-size FFT to minimize the computational complexity. In the second step, a fine estimation is carried out over the selected frequency segment using zoom fast Fourier transform (ZFFT) to improve the estimation accuracy with comparatively low computational complexity. Other OFDM system parameters are also estimated subsequently. Simulation results confirm the improvements of the estimation precision and computational efficiency of the proposed algorithm.

Date Published: 24 May 2011
PDF: 11 pages
Proc. SPIE 8061, Wireless Sensing, Localization, and Processing VI, 80610P (24 May 2011); doi: 10.1117/12.886862

Published in SPIE Proceedings Vol. 8061:
Wireless Sensing, Localization, and Processing VI
Sohail A. Dianat; Michael David Zoltowski, Editor(s)