Carrier frequency offset estimation in OFDM system using ML criterion precoded by turbo codes

Abstract

Frequency offset (FO) introduces severe inter-symbol and inter-carrier interferences which will degrade the performances of orthogonal frequency division multiplexing (OFDM) system more adversely. To reduce the effect of CFO, an estimation algorithm has been proposed in this work which uses a turbo coding at the precoding section and Space time block coding techniques (alamouti code) for transmitting data. The turbo codes solve dilemma of structure and randomness through concatenation and interleaving and it will reduce bit error rate (BER) by error detection and error correction process. A fading channel has been modeled for evaluating the performance of the algorithm. Maximum likelihood (ML) estimation is used for the estimation that does take conditional probability into account. The offset estimation in OFDM is done by best linear unbiased estimator (BLUE) principle yet it has disadvantages that it is already sub optimal in nature and it is applicable only for linear observation models. So, ML technique is an alternative estimation criterion that can use as starting point of most of the statistical estimation and for nonlinear observations. The proposed system has improved BER while other methods supports BER of 25–30 dB and the proposed system shows improvement in spectral efficiency as well as the Power spectral density.