Satellite network: a complete guide to satellite Internet

A satellite network is a wireless communication system that uses one or more artificial satellites to transmit data between different points on Earth or to connect users and terrestrial networks to the Internet (Li, 2025; International Telecommunication Union [ITU], 2022).
Unlike terrestrial networks, where data mainly travels through cables, fiber optics, or radio base stations, in a satellite network part of the path is carried through space using radio signals.
Satellite networks are used for many services, including telecommunications, Internet access, television broadcasting, maritime and aeronautical communications, and connections to areas where building terrestrial infrastructure would be difficult or expensive (ITU, 2022).
A satellite network for Internet access generally includes several elements (Li, 2025):

The satellite should therefore not be thought of as a server that stores the websites requested by the user (Li, 2025).
Its main function is communication and forwarding: it allows data to travel through part of the path between the user and the terrestrial network.
To understand how it works, we can consider a user who uses a satellite connection to open a web page.

The path can be described in simplified form as follows (Li, 2025):

In a modern satellite network, the path does not necessarily always follow:
user → satellite → ground station (Li, 2025).

A satellite constellation is a group of satellites that operate in a coordinated way to provide coverage and communication services over a specific area or, in some cases, over much of the Earth (Li, 2025; ITU, 2022).
Some constellations use Inter-Satellite Links (ISL), which allow one satellite to communicate directly with other satellites in the same constellation (Li, 2025).
Data can therefore travel through part of the constellation before returning to Earth.

Starlink, one of the best-known satellite networks, for example, uses optical laser-based links between its satellites (Starlink, n.d.-c).

Depending on location, link availability, and destination, the network can therefore decide whether to send traffic down to a nearby ground station or forward it through other satellites in the constellation before it reaches the ground.

An important element for understanding satellite networks is orbital altitude, that is, the distance at which a satellite orbits above the Earth’s surface (NASA Earthdata, n.d.).

Orbits can be divided, in simplified form, into three main categories (Allman et al., 1999; NASA Earthdata, n.d.):

GEO satellites travel in a circular orbit above the equator and take the same amount of time as the Earth to complete one revolution.
For this reason, when observed from the Earth’s surface, they appear to remain stationary in the same point in the sky.

This characteristic is very useful in telecommunications (Allman et al., 1999):

However, the great distance from Earth has a disadvantage: the signal must travel tens of thousands of kilometers to reach the satellite and the same distance again to return to the surface (Allman et al., 1999).
Signal propagation therefore introduces a significant delay.
LEO satellites, by contrast, orbit much closer to the Earth’s surface.
The shorter distance reduces the time required for the signal to travel between the Earth and the satellite, resulting in lower delay (Allman et al., 1999).
There is, however, an important consequence: a LEO satellite does not remain stationary above the same area.
From the user’s point of view, the satellite moves quickly across the sky; the terminal must therefore track its movement electronically or periodically switch to another satellite in the constellation.
To provide a continuous connection, constellations are therefore required (Allman et al., 1999).
Starlink follows this principle: instead of relying on a small number of geostationary satellites located very far from Earth, it uses a large constellation of satellites in low Earth orbit, a few hundred kilometers above the surface (Starlink, n.d.-c).
This choice greatly reduces the distance traveled by the signal.
According to specifications published by Starlink, for terrestrial users the service delay is typically in the range of a few tens of milliseconds, although it can vary depending on location, network load, and the path followed by the data (Starlink, n.d.-b).
This situation has some similarities with what happens in mobile networks.
In a mobile network, a moving device can progressively switch from one cell to another.
Something conceptually similar can happen in a LEO satellite network, but in this case it is often the satellite that moves rapidly relative to the user, rather than the mobile phone as in a mobile network.
The terminal must therefore periodically switch communication from one available satellite to another.
This process is called satellite handover.
Satellite handover, or handoff, is the process by which a connection is transferred from one satellite, coverage beam, or link to another while attempting to maintain service continuity (Allman et al., 1999).

In the case of Starlink, the terminal may have several satellites available, and the system dynamically selects the most suitable one (Li, 2025; Starlink, n.d.-c).

Because LEO satellites move quickly across the sky, these changes may occur many times during the normal operation of the connection.

Satellite coverage is not normally described through a single “cell” as in terrestrial cellular networks.

A satellite uses antennas that generate one or more radio beams.
Each beam illuminates a specific area of the Earth’s surface, called a footprint.
A satellite can generate multiple beams and therefore serve several areas simultaneously.

A single satellite can produce multiple beams and reuse its radio resources in different areas (Starlink, n.d.-c).
In LEO constellations, the coverage configuration changes continuously because the satellites move relative to the Earth’s surface.
For this reason, the network must constantly know the positions of satellites, terminals, and ground stations and decide which links to use (Li, 2025).
The antennas used are also different from traditional fixed satellite dishes.
Starlink terminals, for example, use phased-array antennas, consisting of many electronically controlled radiating elements (Starlink, n.d.-c).
Unlike an antenna that must be mechanically pointed toward a satellite, a phased-array antenna can electronically change the direction of the radio beam without continuously rotating the entire antenna (Starlink, n.d.-c).
Starlink satellites also use phased-array antennas and different frequency bands, while links between satellites can use optical laser communications (Starlink, n.d.-c).

As with mobile networks, satellite communications use regulated portions of the electromagnetic spectrum (ITU, 2022).

Available frequencies must be coordinated so that different systems can operate without causing harmful interference (ITU, 2022).
However, connection quality does not depend only on the distance from the satellite.

Because the terminal must communicate with objects in the sky, it is particularly important for it to have a sufficiently clear view of the sky.
Trees, buildings, mountains, or other structures can obstruct the signal path (Allman et al., 1999).

In constellations consisting of many satellites, the system may in some cases switch to another satellite to avoid a temporarily obstructed link, but a location with a severely restricted view of the sky can still compromise the connection.
Weather conditions can also affect propagation. Heavy rain, snow, and hail can attenuate radio signals and temporarily reduce service quality (ITU, 2023).
A satellite connection also shares part of its capacity among multiple users.

The available speed therefore depends not only on terminal technology, but also on the number of users being served, the capacity of the satellites and gateways, geographical location, and network congestion (Li, 2025; Starlink, n.d.-b).
Satellite networks therefore have different advantages and limitations compared with terrestrial infrastructure.
Their main advantage is the ability to reach places where adequate terrestrial connections are not available (ITU, 2022).

This makes satellite networks particularly useful as an access network, that is, as the connection between the user and the operator’s infrastructure (Li, 2025).
However, this does not mean that satellites completely replace the terrestrial Internet.
When traffic returns to Earth, it can continue through points of presence, data centers, routers, fiber-optic backbones, and other traditional infrastructure (Li, 2025).
The satellite network is therefore part of a larger communication system.
Starlink is a significant example of this integration: the user terminal communicates with the satellite constellation, but the constellation is itself connected to a terrestrial network of gateways and points of presence that provide access to the Internet (Li, 2025; Starlink, n.d.-c).

Finally, mobile and satellite networks can be integrated to extend connectivity to areas that are not covered by traditional radio base stations (3rd Generation Partnership Project [3GPP], 2024).

Satellite-to-phone communication, also known as Direct-to-Device (D2D), is a form of communication in which an ordinary mobile phone communicates directly with a satellite without first connecting to a terrestrial radio base station (3GPP, 2024; Starlink, n.d.-a).
The satellite therefore performs, in part, a role similar to that of a radio base station in space.
Starlink Direct to Cell is an example of this type of technology: some Starlink satellites operate as radio base stations in space and can connect directly to compatible mobile phones (Starlink, n.d.-a).

In this case, mobile and satellite networks are no longer completely separate systems, but can become integrated parts of the same communication infrastructure (3GPP, 2024).