SIR Asymptotics in General Network Models

TL;DR

This paper studies the asymptotic behavior of the SIR distribution in diverse wireless network models, revealing tail decay rates and proposing conjectures to predict behavior based on large-scale path loss and interference.

Contribution

It provides a unified analysis of SIR asymptotics across various network models and introduces conjectures linking tail behavior to path loss characteristics.

Findings

Lower tails decay polynomially with the path loss exponent or fading parameter.

Upper tails decay exponentially, except in cellular networks with singular path loss.

Impact of nearest interferer on SIR asymptotics is analyzed.

Abstract

In the performance analyses of wireless networks, asymptotic quantities and properties often pro- vide useful results and insights. The asymptotic analyses become especially important when complete analytical expressions of the performance metrics of interest are not available, which is often the case if one departs from very specific modeling assumptions. In this paper, we consider the asymptotics of the SIR distribution in general wireless network models, including ad hoc and cellular networks, simple and non-simple point processes, and singular and bounded path loss models, for which, in most cases, finding analytical expressions of the complete SIR distribution seems hopeless. We show that the lower tails of the SIR distributions decay polynomially with the order solely determined by the path loss exponent or the fading parameter, while the upper tails decay exponentially, with the exception of cellular networks with singular path loss. In addition, we analyze the impact of the nearest interferer on the asymptotic properties of the SIR distributions, and we formulate three crisp conjectures that -if true- determine the asymptotic behavior in many cases based on the large-scale path loss properties of the desired signal and/or nearest interferer only.