A Stochastic Geometry Based Approach to Tractable 5G RNPO with a New $H$-LOS Model.

We consider a 3D cellular network in which generalized shadowing and radio network planning and optimisation (RNPO) parameters (e.g., antenna height, antenna tilt/azimuth, power biaising ...) are incorporated into the cell-selection model. Using tools from stochastic geometry (SG), we derive an equivalent 2D network in which no shadowing and RNPO parameters are considered. Next, we derive coverage probability for a tractable case-study network, and the regimes where coverage probability is maximized in addition to the interferencelimited one are investigated. An intermediary result is a closedform expression generator encompassing the Q-function based-expression in [1]. Numerical results confirm the accuracy of our approximations.