Date of Award
2011
Degree Type
Thesis
Degree Name
Master of Engineering Science
Program
Electrical and Computer Engineering
Supervisor
Dr. Xianbin Wang
Abstract
Orthogonal frequency division multiplexing (OFDM) has gained worldwide popular ity in broadband wireless communications recently due to its high spectral efficiency and robust performance in multipath fading channels. A growing trend of smart receivers which can support and adapt to multiple OFDM based standards auto matically brings the necessity of identifying different standards by estimating OFDM system parameters without a priori information. Consequently, blind estimation and identification of OFDM system parameters has received considerable research atten tions. Many techniques have been developed for blind estimation of various OFDM parameters, whereas estimation of the sampling frequency is often ignored. Further more, the estimated sampling frequency of an OFDM signal has to be very accurate for data recovery due to the high sensitivity of OFDM signals to sampling clock offset. To address the aforementioned problems, we propose a two-step cyclostation- arity based algorithm with low computational complexity to precisely estimate the sampling frequency of a received oversampled OFDM signal. With this estimated sampling frequency and oversampling ratio, other OFDM system parameters, i.e., the number of subcarriers, symbol duration and cyclic prefix (CP) length can be es timated based on the cyclic property from CP sequentially. In addition, modulation scheme used in the OFDM can be classified based on the higher-order statistics (HOS) of the frequency domain OFDM signal.
All the proposed algorithms are verified by a lab testing system including a vec tor signal generator, a spectrum analyzer and a high speed digitizer. The evaluation results confirm the high precision and efficacy of the proposed algorithm in realistic scenarios.
Recommended Citation
Chen, Qian, "Blind Estimation of OFDM System Parameters for Automatic Signal Identification" (2011). Digitized Theses. 3277.
https://ir.lib.uwo.ca/digitizedtheses/3277