On the Performance of MIMO-SVD Multiplexing Systems in HetNets: A Stochastic Geometry Perspective

In the paper, we study network-wise performance of multistream multi-input multi-output (MIMO) singular value decomposition communications in heterogeneous networks (HetNets). We adopt tools from stochastic geometry to model HetNets through K tiers of interfering base stations (BSs) and evaluate the coverage probability and achievable spectral efficiency (ASE), assuming max-signal-to-interference ratio (SIR) cell association rule. For this model, the main contributions are studying the coverage probability of MIMO multiplexing systems from a communication link's viewpoint; investigating the cross-stream SIR correlation coefficient and highlighting the impacts of path-loss exponent and the number of receive antennas on the growth of it; and obtaining easy-to-compute closed-form approximates of the coverage probability and ASE. The developed expressions explicitly reveal the impact of many system parameters, including the number of tiers, density of BSs, transmission power, and the number of data streams. Simulations are conducted to confirm the accuracy of our analysis. Various important aspects of HetNets with respect to densification, high multiplexing gains, and large antenna arrays are demonstrated. Results showcase the significance of channel state information at the transmitter on the network's performance. With the results of this paper, further investigations and system designs are made possible.