A correlated shadowing model for urban wireless networks
Francois Baccelli, Xinchen Zhang
Abstract
Francois Baccelli, Xinchen Zhang
Abstract
This paper presents an analytically tractable stochastic geometry model for urban wireless networks, where the locations of the nodes and the shadowing are highly correlated and different path loss functions can be applied to line-of-sight (LOS) and non-line-of-sight (NLOS) links. Using a distance-based LOS path loss model and a blockage (shadowing)-based NLOS path loss model, we are able to derive the distribution of the interference observed at a typical location and the joint distribution at different locations. When applied to cellular networks, this model leads to tractable expressions for the coverage probability (SINR distribution). We show that this model captures important features of urban wireless networks, which cannot be analyzed using existing models. The numerical results also suggest that even in the presence of significant penetration loss, ignoring the NLOS interference can lead to erroneous estimations on coverage. They also suggest that allowing users to be associated with NLOS BSs may provide a non-trivial gain on coverage.
OpenAlex reports 91 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper presents an analytically tractable stochastic geometry model for urban wireless networks, where the locations of the nodes and the shadowing are highly correlated and different path loss functions can be applied to line-of-sight (LOS) and non-line-of-sight (NLOS) links. Using a distance-based LOS path loss model and a blockage (shadowing)-based NLOS path loss model, we are able to derive the distribution of the interference observed at a typical location and the joint distribution at different locations. When applied to cellular networks, this model leads to tractable expressions for the coverage probability (SINR distribution). We show that this model captures important features of urban wireless networks, which cannot be analyzed using existing models. The numerical results also suggest that even in the presence of significant penetration loss, ignoring the NLOS interference can lead to erroneous estimations on coverage. They also suggest that allowing users to be associated with NLOS BSs may provide a non-trivial gain on coverage.
Key concepts: Non-line-of-sight propagation, Path loss, Coverage probability, Computer science, Stochastic geometry models of wireless networks, Log-distance path loss model, Interference (communication), Stochastic geometry