Efficient Carrier Frequency Offset Estimation for CO-OFDM Using the Matched Filter Approach

Abstract

This research introduces a unique procedure for estimating carrier frequency offset (CFO) in coherent optical orthogonal frequency division multiplexing (CO-OFDM) systems by applying a unique matching filter (MF). To perform the estimation, the receiver acquires N time-samples of the received CO-OFDM signal, corresponding to an OFDM symbol with N subcarriers. By applying the discrete Fourier transform to these LN samples, we calculate the CO-OFDM symbol spectrum at LN discrete frequency points. The receiver may now determine the hypothesised value that most closely resembles the spectrum of the CO-OFDM symbol received from the channel to produce the maximum likelihood (ML) CFO estimate. The most crucial information in this book is that the DA MF ML method is able to perform full-range (integral plus fractional portion) CFO estimation with good accuracy, whereas the BL MF ML approach solely focuses on fractional CFO estimate and has low computational complexity. To illustrate the estimate error variance for 16-PSK/QAM modulations, we analyse the complexity as well as effectiveness of this MF ML technique also give simulation data.