Stochastic Channel Models for Satellite Mega-Constellations

Brendon McBain, Yi Hong, and Emanuele Viterbo · 2025

Read Stochastic Channel Models for Satellite Mega-Constellations by Brendon McBain, Yi Hong, and Emanuele Viterbo (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-20255.txt.

Introduction

McBain, Hong, and Viterbo build a statistical model of the radio channel between a fast low-orbit satellite and a fixed ground user. Mega-constellations hold thousands of such satellites, and planners need channel models that capture their motion. Their 2025 model scatters satellite positions with a point process whose density varies, and tags each satellite as rising or setting.

From that setup they derive the spread of signal power, delay, and frequency shift. A scattering function then shows how received power spreads across delay and shift. Global traits such as path loss follow from that function. The model's statistics match orbit simulations of the Starlink constellation closely. The authors offer it as a base for designing constellation links.

Catalog record

CREATORBrendon McBain, Yi Hong, and Emanuele Viterbo
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE61 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-20255

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Stochastic Channel Models for Satellite Mega-Constellations Brendon McBain, Yi Hong, and Emanuele Viterbo Abstract A general satellite channel model is proposed for communications between a rapidly moving low Earth orbit (LEO) satellite in a mega-constellation and a stationary user on Earth. The channel uses a non-homogeneous binomial point process (NBPP) for modelling the satellite positions, marked with an ascending/descending binary random variable for modelling the satellite directions. Using the marked NBPP, we derive the probability distributions of power gain, propagation delay, and Doppler shift, resulting in a stochastic signal propagation model for the mega-constellation geometry in isolation of other effects. This forms the basis for our proposed channel model as a randomly time-varying channel. The scattering function of this channel is derived to characterise how the received power is spread in the delay-Doppler domain. Global channel parameters such as path loss and channel spread are analysed in terms of the scattering function. The channel statistics and the global channel parameters closely match realistic orbit simulations of the Starlink constellation. I Introduction In recent years, a growing number of satellites have been launched into low Earth orbit (LEO) to form mega-constellations aimed at expanding global Internet coverage [ 1 ] . However, the high mobility of LEO satellites introduces unique challenges to communication systems [ 2 ] , such as high latency and high Doppler shifts, that are not yet fully understood. Compared to traditional mobile communications, the theoretical understanding of these satellite networks is still in its early stages. As the development of sixth-generation (6G) wireless networks approaches, there is a pressing need to explore optimal design practices for reliable communication in non-terrestrial satellite networks and their integration with terrestrial networks. A key step toward this goal is developing accurate channel models for satellite networks. This paper focuses on stochastic satellite channel models, with an emphasis on capturing the high mobility of satellites in mega-constellations. For modelling the communications channel between a user and a LEO mega-constellation, we must model the satellite positions on their orbit trajectories. Since there are hundreds or thousands of rapidly moving satellites in a mega-constellation, stochastic models like the point process (PP) are appropriate for tractable descriptions of the satellite positions. In [ 3 , 4 ] , Okati et al. proposed the binomial point process (BPP) to model N total satellites as independent and uniformly distributed points on a satellite sphere. Since the BPP did not capture the varying density of satellites at different latitudes, a correction factor was included in the total number of satellites to give the effective number of satellites at different user latitudes. In [ 5 , 6 ] , Okati et al. proposed the non-homogeneous Poisson point process (NPPP) to model the satellites using the probability density of satellites across latitudes, still assuming uniformity across longitudes. This model did not fix the total number of satellites, only the satellite density. In [ 7 ] , the BPP model with the correction factor was extended to include discrete altitudes by distributing the satellites across multiple satellite spheres. These models do not include satellite directions along their orbits, which is the extension that will addressed in this paper. Modelling the satellite positions is sufficient for modelling the power gain due to moving satellites. However, it can only give a channel model of satellite communications after coherent detection, ignoring propagation delay and Doppler effects. This channel model was used to derive coverage probabilities and capacity bounds for mega-constellations [ 3 , 4 , 5 , 6 ] . In [ 8 ] , a survey of thirteen channel models of non-geostationary orbit (NGSO) satellites was conducted, in which more than half of the channel models only considered link attenuation in various scenarios such as multipath or rain attenuation. Link attenuation models are important and complement the mega-constellation models discussed earlier, since they must be combined for a complete channel model. In this survey, it was remarked that almost none of the channel models considered Doppler effects. When Doppler effects between the user and the satellite are considered in the channel model, it is common to use the maximum Doppler shift to parameterise Jakes’ Doppler spectrum [ 9 , 10 ] . However, using this Doppler spectrum is a heuristic since it is not derived from the satellite mega-constellation geometry. In contrast, the approach in this paper will characterise the Doppler shift based on a geometric model of mega-constellations. To highlight the novelty in this approach, the stochastic channel models surveyed in [ 8 ] are classified in Table I in terms of if the models are based on mega-constellation geometry or link attenuation, and if they include Doppler shift. Model type Mega-constellation modelling Link modelling Doppler shift inclusion BPP [ 3 , 4 , 7 ] ✓ ✓ ✗ NPPP [ 5 , 6 ] ✓ ✓ ✗ TDL [ 10 ] ✗ ✓ ✓ Elevation-dependent [ 11 , 12 ] ✗ ✓ ✗ Small-scale fading [ 13 ] ✗ ✓ ✗ Rain fading [ 14 ] ✗ ✓ ✗ Marked NBPP (proposed) ✓ ✓ ✓ TABLE I : A classification of the stochastic satellite channel models surveyed in [ 8 ] . In [ 15 ] , the Doppler shift was characterised for a stationary user and a moving LEO satellite. This characterisation considered the time-dependent movement of the satellite on its orbit under the assumption of a mega-constellation geometry with circular orbits. On the other hand, stochastic Doppler characterisations using stochastic geometry have been studied in just a couple of scenarios. In [ 16 ] , the Doppler distribution was derived for a single LEO satellite in a fixed position transmitting to a random user in a cell on Earth. In [ 17 ] , the Doppler distribution was derived for a LEO satellite transmitting to an unmanned aerial vehicle (UAV) randomly moving in a swarm. In [ 18 ] , a closed-form distribution of a Doppler upper bound was derived for the scenario of a random LEO satellite from a mega-constellation transmitting to a stationary user. This scenario is the focus of this paper in the context of stochastic channel modelling, which will be shown to very accurately characterise the true Doppler shift distribution. In addition, the relationship between Doppler shift, propagation delay, and channel gain (inverse of free space path loss) is studied, which has not been considered in the literature for a stochastic model of the mega-constellation geometry. In satellite communications, many useful techniques exist to compensate for propagation delay and Doppler shift. Nonetheless, understanding the channel before compensation can aid the design and testing of compensation techniques. Moreover, there are many possible approaches to compensation and they are highly dependent on the configuration of the mega-constellation network, e.g., the particular inter-satellite handover strategy alone can greatly influence delay and Doppler shift [ 19 ] . In addition, Doppler shift compensation is known to be challenging in cooperative satellite communications [ 20 , 21 ] . This paper does not include compensation so that the scope of the results are not limited. However, the methods developed are sufficiently general to support the inclusion of an arbitrary compensation technique. For this reason, the channel model that will be

Read the complete text