Beyond Starlink: Is Satellite Really the Answer to Communications Black Spots?
Satellite broadband has changed what is possible in remote communications.
Low-Earth-orbit services such as Starlink can deliver useful internet access to homes, businesses, vessels and remote installations far beyond the practical reach of fibre or conventional broadband. In the right location, satellite can turn an isolated site into a connected one remarkably quickly.
But satellite coverage is sometimes presented as though it has eliminated the communications black-spot problem.
It has not.
Satellite can overcome certain geographical gaps, but it introduces its own physical, operational and commercial dependencies. It may be an excellent source of internet backhaul while still being the wrong primary communications system for a battery-powered sensor, an underground plant room, a moving asset or a safety-critical alarm.
The important question is therefore not:
“Can we see a satellite?”
It is:
“Can the data reach the right person or system reliably, economically and quickly enough to achieve its purpose?”
What satellite broadband does well
Traditional geostationary communications satellites orbit approximately 35,786 kilometres above the Earth. This enables a satellite to cover a very large area, but the distance creates noticeable latency.
Low-Earth-orbit—or LEO—constellations operate much closer to Earth. This can reduce latency and increase bandwidth, making services such as web access, video calls and cloud applications practical in locations previously dependent on slow or expensive connections.
Satellite broadband is particularly valuable for:
isolated homes and workplaces;
rural businesses;
temporary construction sites;
ships and offshore operations;
emergency recovery;
remote offices;
geographically dispersed infrastructure;
locations where installing terrestrial broadband is uneconomic.
These are important capabilities. Satellite should be regarded as a major addition to the communications landscape.
But an additional communications option is not the same as universal, uninterrupted coverage.
Satellite still needs a clear route to the sky
A satellite terminal must maintain a radio path to satellites moving across its visible sky.
Starlink itself advises users to install terminals with a clear field of view. Trees, branches, poles, roofs and other obstructions can cause service interruptions. Moderate or heavy rain, snow and hail can also produce degradation or momentary dropouts.
This creates an immediate limitation: many communications black spots exist precisely because radio signals are being blocked.
Satellite can perform well in an open rural field while struggling or becoming unusable in:
basements;
tunnels;
mines;
underground car parks;
steel-framed factories;
enclosed plant rooms;
urban streets surrounded by tall buildings;
dense forests;
valleys and cuttings;
the interior of vehicles or containers;
installations beneath machinery or infrastructure.
The terrestrial mobile signal may be absent because the location is physically difficult. That same location may also have no adequate view of the sky.
Satellite does not abolish radio physics.
Coverage is not the same as continuity
A service may be geographically available while still experiencing short interruptions.
For ordinary internet use, a brief dropout may pass unnoticed because applications buffer data, retry connections or resume automatically. For industrial and sensor applications, however, the significance depends on what was supposed to happen during the interruption.
A five-second break could be irrelevant to an hourly environmental report. The same five seconds could matter greatly if the missing data contained:
a sudden pressure change;
an equipment trip;
a lone-worker alarm;
a temperature excursion;
a security event;
a fire or gas warning;
a control instruction;
a rapidly changing energy condition.
Suppose a sensor records ten readings every second. A 15-second interruption affects:
[10 \times 15 = 150\text{ readings}]
Across 200 sensors, the same event affects:
[200 \times 150 = 30{,}000\text{ readings}]
Those readings are not necessarily lost—well-designed equipment may store and retransmit them—but they are unavailable to the remote system during the interruption.
For operational data, event time and delivery time are not the same thing.
Satellite may restore eventual delivery. It cannot make delayed information contemporaneous again.
The power problem
Satellite broadband terminals, routers and associated equipment require a dependable electrical supply.
That is reasonable for a building, vessel or permanent site with mains power. It can be problematic for a small sensor expected to operate for months or years from a battery.
A pressure, temperature or environmental sensor may need to transmit only a few bytes at a time. Maintaining a broadband satellite terminal solely to carry these small messages can be disproportionate in energy, equipment and cost.
There is a fundamental difference between:
providing broadband to a site; and
transmitting a small, occasional sensor message.
A broadband terminal is designed to move substantial amounts of data. Many IoT applications require very little bandwidth but demand long battery life, predictable delivery and low operating cost.
The most powerful connection is not automatically the most suitable one.
Installation and physical exposure
A satellite terminal usually needs to be mounted where it can see the sky. This can place it on a roof, mast, vehicle exterior or exposed structure.
That introduces practical considerations:
installation labour;
cabling;
mounting and alignment;
wind loading;
lightning protection;
accidental or deliberate damage;
snow, dirt or foliage accumulation;
access for inspection and maintenance;
weather sealing;
power continuity.
Starlink notes that new foliage, material deposited on the terminal and snow accumulation can obstruct service.
Satellite may avoid the cost of installing kilometres of cable, but it does not remove the need for properly engineered equipment at the site.
Shared capacity and variable performance
A satellite system has finite radio spectrum and network capacity. Users within an area share that capacity, just as users share terrestrial mobile infrastructure.
Performance can therefore vary with:
local demand;
the number of active terminals;
network management;
satellite availability;
ground infrastructure;
weather;
obstructions;
the service plan;
movement of the terminal;
traffic routing.
For general internet access, variability may be acceptable. For operational technology, the relevant questions are more demanding:
What is the worst-case delivery time?
What happens when the connection disappears?
Is the message retained?
Is delivery confirmed?
Can an alarm still be shown locally?
Can the system operate without its remote platform?
How quickly does it recover?
Is the interruption visible to the user?
Average download speed does not answer these questions.
Satellite creates another external dependency
Satellite removes dependence on nearby terrestrial infrastructure, but it does not remove infrastructure altogether.
The operating chain may still include:
[\text{Sensor} \rightarrow \text{Local Network} \rightarrow \text{Gateway} \rightarrow \text{Satellite Terminal} \rightarrow \text{Satellite Network} \rightarrow \text{Internet} \rightarrow \text{Server} \rightarrow \text{Application}]
The chain is impressive, but it remains a chain.
Power failure, damaged cabling, terminal obstruction, equipment failure, service suspension, network congestion, software faults or problems with the remote platform can still interrupt the outcome.
If every one of seven required stages had an availability of 99.5%, the simplified end-to-end availability would be:
[0.995^7 = 0.9655]
or approximately 96.6%.
This example assumes independent failures and identical availability, which real systems rarely have. Its purpose is to demonstrate the effect of accumulating dependencies: highly reliable components do not automatically create a highly reliable end-to-end service.
Satellite broadband and satellite IoT are different
Not every satellite system is a broadband system.
Specialised satellite IoT services can transmit small messages using lower-power equipment. They may be better suited to remote assets, agriculture, maritime operations, environmental monitoring and logistics.
The mobile-communications industry is also developing standardised non-terrestrial networks. 3GPP Release 17 introduced support for non-terrestrial 5G and IoT communications, including NB-IoT over satellite, with further development continuing in later releases.
These services can be more appropriate than full broadband where the requirement is a small telemetry message rather than a permanent high-speed connection.
They still face fundamental constraints:
access to the sky;
antenna orientation and efficiency;
transmission power;
service availability;
message latency;
subscription costs;
network dependency;
limited or scheduled transmission opportunities.
Satellite IoT can be extremely useful. It should be selected because it matches the application—not simply because the word “satellite” implies universal coverage.
What may work better?
No single communications technology is superior in every situation. Different systems solve different parts of the black-spot problem.
1. Multi-network cellular
A device restricted to one mobile operator may enter a black spot even where another operator has coverage.
Multi-network SIMs, roaming arrangements and eSIM-based connectivity can allow compatible devices to use more than one available network. This can materially improve resilience along transport routes and across dispersed installations.
Mobile IoT technologies such as LTE-M and NB-IoT are designed for connected devices requiring lower power and less bandwidth than a conventional broadband connection. GSMA deployment guidance treats NB-IoT and LTE-M as complementary low-power wide-area technologies and emphasises interoperability and roaming.
Multi-network cellular is attractive where:
at least one operator usually has coverage;
devices move between coverage areas;
modest data volumes are required;
national or international roaming is available;
the application can tolerate occasional reconnection.
Its limitation is straightforward: if no operator reaches the location, having access to more operators does not create a signal.
2. Private LTE or 5G
A business can deploy its own cellular infrastructure across a factory, port, campus, warehouse, mine or industrial site.
A private network can provide greater control over coverage, capacity, security and service quality than a public mobile network. It can be engineered around the site rather than around mass-market consumer demand.
Private cellular can be appropriate where:
many devices require coverage;
traffic volumes are substantial;
mobility must be supported;
the site requires controlled access;
low latency is important;
the organisation can justify professional deployment and management.
Its disadvantages include cost, specialist expertise, spectrum arrangements and continuing infrastructure requirements. It is generally more suitable for a substantial site than for a handful of simple sensors.
3. Low-power wide-area radio
Low-power wide-area networks are designed to carry small amounts of data across relatively long distances using modest energy.
LoRaWAN, for example, connects battery-operated devices through gateways using a star-of-stars architecture. A sensor may be heard by more than one gateway, providing useful redundancy. The technology is designed around low-power IoT rather than high-bandwidth internet access.
LPWAN systems can be well suited to:
metering;
environmental monitoring;
agriculture;
asset status;
temperature and humidity;
energy data;
alarms;
widely dispersed low-data-rate sensors.
However, conventional gateway-based LPWAN may still depend on the gateway having backhaul to a server. If the gateway loses internet access and cannot operate or store data locally, the black spot has merely moved from the sensor to the gateway.
4. Distributed radio networks
In a distributed network, devices can pass information through neighbouring devices rather than every device requiring a direct path to a central mast or gateway.
This can be particularly useful where terrain, buildings or distance prevent a single radio from reaching the destination.
A route might look like:
[A \rightarrow B \rightarrow C \rightarrow D]
If an alternative path exists, the network may reroute:
[A \rightarrow E \rightarrow F \rightarrow D]
This ability to form alternative paths can create resilience in industrial estates, farms, towns, campuses, utility networks and emergency deployments.
Open field-area networking standards are already moving in this direction. Wi-SUN, for example, is designed for secure, interoperable, self-forming and self-healing mesh networks used in utilities and smart-city infrastructure. Its Field Area Network specification was ratified as ISO/IEC/IEEE 32857:2026.
ETSI’s DECT-2020 standard also supports autonomous, non-cellular networking in which devices can function as nodes and routers without dependence on a commercial mobile operator.
Distributed networking has its own constraints. Performance depends upon:
node density;
radio range;
terrain and obstructions;
available spectrum;
message size;
duty-cycle restrictions;
routing efficiency;
battery capacity;
network security;
the presence of a viable route.
It is not unlimited-bandwidth internet. Its strength lies in moving essential data through a network designed around the local environment.
5. Store-and-forward communications
Some black spots cannot be eliminated continuously. In those circumstances, the system must be designed to survive interruption.
Store-and-forward operation allows a device to:
create and time-stamp a message;
store it securely;
attempt delivery;
retain it if no route is available;
forward it when a connection or neighbouring node becomes available;
confirm successful receipt;
prevent duplication.
This is the principle behind delay- and disruption-tolerant networking. The IETF’s DTN architecture uses persistent storage and forwarding to overcome communication interruptions between heterogeneous networks.
Store-and-forward cannot make delayed data arrive in real time. It can, however, prevent a temporary communications gap from becoming permanent data loss.
For metering, logistics history, environmental records and non-urgent sensor information, that distinction is enormously valuable.
6. Local processing and local display
Sometimes the most dependable way to overcome a remote communications interruption is to avoid making the immediate outcome dependent upon remote communications.
A sensor value can be:
received locally;
validated locally;
displayed locally;
assessed against local thresholds;
stored locally;
acted upon locally;
forwarded externally when a suitable route becomes available.
This does not eliminate the need for external communication. It separates two requirements that are too often treated as one:
Must the information be available here and now?
Must the information also be available somewhere else?
The first requirement may be fulfilled locally even when the second is temporarily impossible.
The mathematics of multiple paths
Communications resilience improves when a system has genuinely independent routes.
If one path has 95% availability, its unavailability is:
[1-0.95=0.05]
If two independent paths each have 95% availability, both are unavailable simultaneously with probability:
[0.05 \times 0.05 = 0.0025]
The combined theoretical availability becomes:
[1-0.0025=99.75%]
Adding another route can improve the figure further.
But independence is critical. A satellite terminal and mobile gateway powered by the same failed supply are not independent. Two cloud applications using the same backhaul are not independent. Multiple radio links mounted behind the same obstruction may fail together.
Redundancy is meaningful only when the failure modes are different.
Comparison at a glance
Communications system | Principal strength | Principal weakness | Best suited to |
LEO satellite broadband | High-speed access in remote open locations | Requires sky view, power and external service | Remote premises, vessels and site backhaul |
Satellite IoT | Wide-area transmission of small messages | Sky access, latency and service cost | Very remote telemetry and moving assets |
Multi-network cellular | Uses existing national infrastructure | Cannot overcome a total mobile not-spot | Vehicles, mobile workers and dispersed assets |
Private LTE/5G | Controlled site-wide capacity and coverage | Cost and deployment complexity | Factories, ports, campuses and large sites |
Gateway-based LPWAN | Long-range, low-power sensor communication | Gateway and backhaul may remain dependencies | Metering, agriculture and environmental sensing |
Distributed radio network | Alternative local paths and reduced infrastructure dependency | Requires suitable node density and routing | Utilities, estates, communities and field operations |
Store-and-forward | Preserves data through interruptions | Does not provide instantaneous remote delivery | Logistics records, metering and delayed telemetry |
Local processing/display | Immediate information without remote connectivity | Remote visibility still requires a later route | Safety, equipment, energy and situational awareness |
The wrong question produces the wrong system
Asking “Which technology has the greatest range?” is not enough.
A more useful assessment asks:
How much data must be transmitted?
How quickly must it arrive?
What happens if it arrives late?
Must the device operate from a battery?
Is the sensor indoors, underground or moving?
Is local information sufficient during an outage?
How many independent communication routes exist?
Can data be stored securely?
Can messages move through neighbouring devices?
Is continuing operation dependent upon a server or cloud platform?
What is the cost of infrastructure over the full product life?
Who controls the data and the communications path?
The best answer may be satellite. It may be cellular, private radio, low-power wide-area networking, distributed communications or local processing.
Increasingly, the strongest answer will be a carefully controlled combination.
Tigertek’s view: continuity before connectivity
Tigertek does not regard internet access as the same thing as operational certainty.
Satellite can provide valuable connectivity, particularly as remote backhaul. But a dependable sensor architecture should not assume that any single external network—satellite, mobile, Wi-Fi or wired internet—will always be present.
The essential data should remain useful as close as possible to the place where it is created.
That means thinking beyond coverage maps and headline speeds. It means considering local visibility, data integrity, storage, alternative paths, power consumption, failure modes and the commercial consequence of interruption.
The future is unlikely to be satellite or terrestrial communications.
It will be an intelligent combination of local operation and appropriately selected communications routes:
[\text{Sense locally} \rightarrow \text{Understand locally} \rightarrow \text{Communicate when required}]
The objective is not permanent dependence on a particular network.
The objective is continuity of the outcome.
