Channel Estimation and Data Equalization in Frequency-Selective MIMO Systems with One-Bit Quantization
This paper addresses channel estimation and data equalization on frequency-selective 1-bit quantized Multiple Input-Multiple Output (MIMO) systems. No joint processing or Channel State Information is assumed at the transmitter, and therefore our findings are also applicable to the uplink of Multi-User MIMO systems. System models for both Orthogonal Division Frequency Multiplexing (OFDM) and single-carrier schemes are developed. A Cramér-Rao Lower Bound for the estimation problems is derived. The two nonlinear algorithms Expectation Maximization (EM) and Generalized Approximate Message Passing (GAMP) are adapted to the problems, and a linear method based on the Bussgang theorem is proposed. In the OFDM case, the linear method enables subcarrier-wise estimation, greatly reducing computational complexity. Simulations are carried out to compare the algorithms with different settings. The results turn out to be close to the Cramér-Rao bound in the low Signal to Noise Ratio (SNR) region. The OFDM setting is more suitable for the nonlinear algorithms, and that the linear methods incur a performance loss with respect to the nonlinear approaches. In the relevant low and medium SNR regions, the loss amounts to 2-3 dB and might well be justified in exchange for the reduced computational effort, especially in Massive MIMO settings.