Stochastic Geometry-based Modelling of Mobile UAV Relay Networks under Realistic Fading

TL;DR

This paper models UAV relay networks using stochastic geometry, analyzing coverage probability considering realistic fading, UAV mobility, and different link conditions, supported by simulations and theoretical derivations.

Contribution

It introduces a stochastic geometry framework for UAV relay networks incorporating mobility, realistic fading, and coverage analysis, providing new insights into UAV deployment strategies.

Findings

UAV altitude and density significantly impact coverage probability.

Mobility schemes can enhance network coverage and reliability.

Stochastic geometry effectively models complex UAV relay network behaviors.

Abstract

We consider a relay network based on Unmanned Aerial Vehicles (UAV). Terrestrial Base Stations (TBS) and UAV Relay Nodes (RN) are modelled using two Homogeneous Poisson Point Processes (HPPP). UAVs can hover at a fixed position or move following specific displacement schemes. The Coverage Probability (CP) of a typical user equipment (UE) is derived, either when it communicates via a direct link (from the TBS to the UE) or via a relay link (from the TBS to the UE through a UAV RN). Every link can be in Line-of-Sight (LoS) or Non Line-of-Sight (NLoS), and suffers from Rician fading with distance-dependent parameters. This coverage is calculated by means of both stochastic geometry (SG) and Monte-Carlo (MC) simulations. The benefits of the use of UAV as RNs are analysed depending on their altitude, density, and mobility scheme.