A complexity efficient equalization technique for MIMO-constant envelope modulation

Abstract

MIMO-Constant Envelope Modulation, MIMO-CEM, with Low Resolution ADC receiver (1-bit ADC in the default operation) is considered as one of the alternative candidates to the well-known MIMO-OFDM, suitable for wireless backhaul networks. Because MIMO-CEM receiver is based upon low resolution ADC sampled at the Intermediate Frequency (IF) band, the currently used MIMO-CEM equalizer is too complicated. The equalizer high complexity comes from its design; in which IF based Maximum Likelihood Sequence Estimator (MLSE) is utilized. In order to efficiently estimate the transmitted sequence, the MLSE equalizer replicates the whole IF MIMO-CEM transceiver, which results in too complicated MIMO-CEM equalizer. In this paper, low complexity equalization technique for MIMO-CEM is proposed. In the proposed method, a baseband approximation model for the 1-bit ADC MIMO-CEM is investigated. Through using this model, a baseband based MIMO-CEM MLSE equalizer can be constructed. This baseband equalizer will nominate number of candidate sequences, which are further refined by the IF based MLSE equalizer to accurately estimate the transmitted sequence. Hence, the number of equalization processes required by the high complexity IF based MLSE are greatly reduced, which powerfully contributes in reducing the complexity of MIMO-CEM equalization. The effectiveness of the proposed estimator is proven under different scenarios using Minimum Shift Keying (MSK) and Gaussian MSK (GMSK) modulations.