OFDM impairment mitigation techniques

In this thesis, we propose novel techniques to improve the performance of OFDM systems. We present a simple adaptive modulation technique to mitigate the nonlinear distortion effects of OFDM signals. Based on an estimation of the nonlinearity of the HPA/channel, for each OFDM symbol, a calculation is done at the transmitter side which identifies the subcarries with high distortion and correpsondingly reduces the modulation level on those subcarriers. This procedure is repeated until the nonlinear distortion is below a predetermined threshold. This technique is shown to improve the BER performance considerably while the reduction in data rate is small. The data rate is reduced by 4% for a system with 64 subcarriers and 16 QAM as primary and 4 QAM as secondary modulation levels. The tone reservation technique used in conventional PAPR reduction is suitably modified to provide a simple solution to reduce the average power requirement in intenstiy modulated optical OFDM systems. With two reserved subcarriers the reduction in power is approximately 2 dB for 16 QAM modulation with 64 subcarriers and 1dB for 256 subcarriers. We also describe techniques to improve the BER performance of grouped linear constellation precoding (GLCP) OFDM which is used to achieve frequency diversity. We present an Adaptive Weighting (AW) technique in which during MLSE, the distance of the possible constellation points from the received symbols are weighted according to the SNR of each subcarrier before the sequence with the minimum distance is chosen. We also analyze a sub optimum iterative decoding algorithm which improves the performance of an initial zero forcing detection iteratively on a symbol by symbol basis. Both techniques improve the performance of the GLCP-OFDM system considerably at the expense of increased complexity.