Channel estimation in massive MIMO systems using 1-bit quantization

Massive MIMO plays an important role for future cellular networks since the large number of antennas deployed at the base station (BS) is capable of increasing the spectral efficiency and the amount of usable spectrum. Using 1-bit analog-to-digital converters can drastically reduce the resulting complexity and power consumption. Therefore, we investigate the channel estimation in 1-bit massive MIMO, where several single-antenna mobile stations (MSs) transmit training sequences to the BS, whose antennas acquire only 1-bit measurements. The channels between the MSs and the BS antennas are described by their impulse responses. In particular, we consider sparse channel impulse responses having only a few non-zero taps. By combining the Expectation-Maximization algorithm for Maximum A Posteriori estimation with the sparse recovery method Iterative Hard Thresholding, we exploit the a priori knowledge of this sparsity and take the 1-bit quantization into account. Since the resulting channel estimation methods combine a good channel estimation performance demonstrated by numerical results with a small computational complexity, they are promising methods for channel estimation in 1-bit massive MIMO.