Modeling and analysis of ultra-dense networks under LOS and NLOS benchmark association scenarios using a stochastic-geometry approach

Abstract

Ultra-dense networks (UDNs) have become a key enabler for meeting the capacity demands of 5G systems. However, their performance is strongly shaped by blockage effects and the resulting mixture of line-of-sight (LOS) and non-line-of-sight (NLOS) propagation. This paper develops a tractable analytical framework for evaluating downlink coverage probability in UDNs under closest-LOS and closest-NLOS benchmark association scenarios, which are used to isolate the effects of LOS- and NLOS-dominated serving conditions. Base stations are modeled using a homogeneous Poisson point process with distance-dependent LOS probability, and a bounded path-loss model is applied to ensure physical realism at short distances. Closed-form expressions for the probability density functions of the nearest LOS and NLOS base stations are derived, enabling analytical evaluation of signal-to-interference-plus-noise ratio coverage under both propagation states. The accuracy of the framework is validated through extensive Monte Carlo simulations, which show excellent agreement with theoretical predictions. Numerical results demonstrate the distinct performance behavior of LOS and NLOS links under densification and quantify their combined impact on overall network reliability. The findings provide practical insights into blockage-aware UDN planning and the performance impact of LOS- and NLOS-dominated serving conditions.