Reduced Complexity Equalization of General Multicarrier System in Time and Frequency Domain

Joint equalization of general multicarrier transmission systems is investigated. The optimal two-dimensional maximum likelihood sequence estimation (MLSE) receiver uses delay spread in time and spectral overlap in frequency as diversity to jointly equalize all carriers. The optimal equalization is computationally intensive, therefore different suboptimal equalization schemes are studied. Simulation results show that reduced complexity schemes in frequency domain (reduced group size of carriers, sequential detection and variants thereof) perform better than conventional one-dimensional receiver, however computational overhead and error propagation are major problems. Reduced complexity schemes in time domain truncate the length of ISI and ICI channel impulse response. It is concluded that latter provide better performance/complexity tradeoff than former.