Coverage and rate analysis for non-uniform millimeter-wave heterogeneous cellular network

Millimeter wave (mmWave) is a promising approach for the fifth generation cellular networks. It has a large available bandwidth and high gain antennas, which can offer interference isolation and overcome high frequency-dependent path loss. In this paper, we study the non-uniform heterogeneous mmWave network. Non-uniform heterogeneous networks are more realistic in practical scenarios than traditional independent homogeneous Poisson point process (PPP) models. We derive the signal-to-noise-plus-interference ratio (SINR) and rate coverage probabilities for a two-tier non-uniform millimeter-wave heterogeneous cellular network, where the macrocell base stations (MBSs) are deployed as a homogeneous PPP and the picocell base stations (PBSs) are modeled as a Poisson hole process (PHP), dependent on the MBSs. Using tools from stochastic geometry, we derive the analytical results for the SINR and rate coverage probabilities. The simulation results validate the analytical expressions. Furthermore, we find that there exists an optimum density of the PBS that achieves the best coverage probability and the change rule with different radii of the exclusion region. Finally, we show that as expected, mmWave outperforms microWave cellular network in terms of rate coverage probability for this system.