Coverage in Terahertz Cellular Networks with Imperfect Beam Alignment

We develop a novel stochastic geometry framework to quantify the SINR coverage probability of a Terahertz (THz) cellular network. THz frequencies will require highly directional beams, leading to inevitably imperfect beam alignment. We derive a tractable and accurate semi-closed form lower bound for the coverage probability of a typical user in the network, introducing a novel approach of characterizing non line-of-sight (NLOS) links as equivalent LoS links using a non-homogeneous Poisson Point Process. We use the coverage bound to investigate the SINR scaling trends with base station density and array directivity at the base station and user. Dense base station deployments are required to achieve sufficient coverage and our analysis exposes a tradeoff between directivity (array gain) and loss in SINR due to misalignment.