Underlay Radar-Massive MIMO Spectrum Sharing: Modeling Fundamentals and Performance Analysis

In this work, we study underlay radar-massive MIMO cellular coexistence in LoS/near-LoS channels, where both systems have 3D beamforming capabilities. Using mathematical tools from stochastic geometry, derive an upper bound on the average interference power at radar due to massive downlink under worst-case `cell-edge beamforming' conditions. To overcome technical challenges imposed by asymmetric and arbitrarily large cells, devise a novel construction which each Poisson Voronoi (PV) cell is bounded its circumcircle effect of random shapes interference. Since model intractable for further analysis correlation between adjacent PV cells' sizes, propose tractable nominal model, as circular disk with area equal typical cell. We quantify gap these two models show that tight realistic deployment parameters. also compare them more practical but MU-MIMO scheduling our same trends do not deviate significantly scheduler models. Under characterize distribution using dominant interferer approximation deriving equi-interference contour expression when receiver uses beamforming. Finally, use expressions radar's spatial probability false alarm/detection quasi-static target tracking scenario. Our results reveal useful function parameters (BS density, exclusion zone radius, antenna height, transmit BS, etc.).