Massive Distributed Antenna Systems: Channel Estimation and Signal Detection

This article addresses the uplink information transmission in the massive distributed antenna (mDA) systems, where the density of antenna elements (AEs) can be much higher than the density of users. We conceive that such mDA systems are user-centric and provide various advantages for practical operations, and that there are evident challenges in the signal detection, which is very different from that in the conventional cellular systems with co-located antennas at base-stations (BS). Therefore in this article, based on the theory of stochastic geometry, some insightful analysis for the achievable performance of mDA systems is first provided, to show that local detection can be more practical than global detection. Then, by exploiting the sparsity of the channel matrices describing mDA systems, a channel estimator is proposed based on the orthogonal matching pursuit (OMP) algorithm. We will show that the OMP-based channel estimation can not only provide channel state information (CSI), but also assist to select the users to be served by an AE. Furthermore, benefiting from the above-mentioned sparsity, we propose a distributed low-complexity message passing algorithm (MPA)-assisted multiuser detector. Finally, we investigate the performance of the mDA systems employing the proposed detection schemes and with various system settings. Our studies show that the OMP-based channel estimation and MPA-assisted detector are capable of achieving a good trade-off among BER performance, complexity and resource usage.