Massive MIMO Systems Characterization Based on Spatial Channel Correlation

Massive multiple-input multiple-output (MIMO) is a breakthrough technology advancing huge network capacities in multiuser scenarios in which a base station (BS), with a large number of antennas, simultaneously serves multiple users in the same time-frequency resources. We study the massive MIMO channel behavior through the channel matrix singular value spread (SVS), users' channels correlation, and sum-rate. As the inter-user spacing or the number of BS antennas increase, the sum-rate, SVS and inter-user correlation enhance and the users' channels can be easily separated. Concerning the duplexing transmission schemes, in time division duplexing (TDD) mode, the entire bandwidth for both uplink and downlink is utilized and the propagation channel reciprocity can be used where the amount of resources needed for pilots only depends on the number of simultaneously served terminals. However, in frequency division duplexing (FDD) systems, which are widely deployed because of their existing spectrum assignments, the uplink and downlink channels are at different frequencies and thus are not reciprocal where a considerable amount of feedback overhead is required. Our central objective is to exploit the directional spatial correlation for the uplink and downlink channels, based on the structure of their multipath clusters, to estimate the downlink channel for an FDD system. The proposed method uses the spatial correlation between the uplink and downlink where the clusters are deduced from the uplink channel. Then, the phase of the signal arriving at the BS is modified to construct the signal departing from the BS and hence the downlink channel can be estimated.