The Best-cell ISR Modeling for Network Planning

The interest in finding an analytical expression for the downlink signal-to-interference ratio (SIR), a key quantity for network planning, has never dwindled; especially with the rapid deployment of 5G networks, dynamic cell dimensioning is deemed necessary for energy saving as well as for multi-tenancy provisions. In this paper, we propose an alternative approach to the much discussed Poisson point process (PPP) network model: Taking into account the cell selection that is subject to path loss with log-normal (LN) shadowing, we model both other-cell interference (OCIF) and interference-to-signal ratio (ISR) for the best cell of a multi-carrier network. The results are tractable and cell-geometry independent, hence applicable to any grid network of any co-channel BS density. We demonstrate that the weighted spatial average of the ISR’s cumulative distribution function (cdf) with the location probability distribution provides a more accurate estimation for the cell coverage probability. We also show that the LN shadowing, which is unaccounted for with the PPP network model, has notable effect on the variance of the cell coverage probability. Additionally, we employ the weighted spatial average of OCIF with the location probability distribution to evaluate the appropriate BS transmission power for various BS densities.