A Dominant Interferer plus Mean Field-based Approximation for SINR Meta Distribution in Wireless Networks

TL;DR

This paper introduces a new dominant interferer plus mean-field approximation method for calculating the SINR meta distribution in wireless networks, simplifying analysis while maintaining accuracy across various network models.

Contribution

It presents a novel approximation technique that reduces complexity by avoiding joint distance computations, applicable to multiple point process models including the first derivation for Poisson line Cox processes.

Findings

The approximation matches well with Monte Carlo simulations.

It accurately predicts SINR meta distribution across different network models.

First derivation of SINR meta distribution for Poisson line Cox processes.

Abstract

This paper proposes a novel approach for computing the meta distribution of the signal-to-interference-plus-noise ratio (SINR) for the downlink transmission in a wireless network with Rayleigh fading. The novel approach relies on an approximation mix of exact and mean-field analysis of interference (dominant interferer-based approximation) to reduce the complexity of analysis and enhance tractability. In particular, the proposed approximation omits the need to compute the first or the second moment of the SINR that is used in the beta approximation typically adopted in the literature but requires of computing the joint distance distributions. We first derive the proposed approximation based on a Poisson point process (PPP) network with a standard path-loss and Rayleigh fading and then illustrate its accuracy and operability in another four widely used point processes: Poisson bipolar network, Mat\'{e}rn cluster process (MCP), $K$-tier PPP and Poisson line Cox process (PLCP). Specifically, we obtain the SINR meta distribution for PLCP networks for the first time. Even though the proposed approximation looks simple but it shows good matching in comparison to the popular beta approximation as well as the Monte-Carlo simulations, which opens the door to adopting this approximation in more advanced network architectures.