Stratospheric Internet Could Connect the Billions Still Offline

While satellite mega-constellations dominate the headlines, engineers are placing their hopes on another layer of the atmosphere: the stratosphere. They believe it could connect billions of people who remain offline more affordably and reliably than space-based systems can achieve on their own.

Nearly a quarter of humanity remains offline

By 2026, Earth’s orbit will be crowded with spacecraft. Starlink is expected to operate around 10,000 satellites, while OneWeb is targeting roughly 650. Telecommunications marketing teams readily promise “global coverage”.

The picture on the ground, however, is far less straightforward.

The ITU’s “Facts and Figures 2025” report estimates that about 2.2 billion people still do not have a usable internet connection, many of them in rural or isolated locations. Nearly one in four people is either entirely unconnected or dependent on links that are painfully slow and unreliable.

Even with thousands of satellites overhead, connectivity gaps remain stubborn, especially in remote and low-income regions.

Satellite networks face three fundamental constraints:

• Capacity constraints: Operating hundreds of kilometres overhead, each satellite has to serve an enormous area. As more users connect, speeds can fall dramatically.
• Cost and complexity: Constructing and maintaining a dense low-Earth orbit constellation capable of reaching every part of the globe is technically complex and exceptionally costly.
• User pricing: Hardware and subscription prices remain well beyond the means of many households in developing countries.

Telecommunications companies are therefore looking towards a lower, less costly layer of the sky to close these gaps.

Stratospheric internet: the layer between Earth and space

This developing option is called stratospheric internet and relies on HAPS, or High Altitude Platform Stations. These can be long-endurance aircraft, balloons or airships flying approximately 18 to 25 kilometres above sea level. They operate well above commercial airliners, yet far below satellites, which orbit at about 500 kilometres or more.

HAPS can include several types of vehicle:

• Helium airships
• Super-pressure balloons
• Solar-powered drones or gliders
• Uncrewed fixed-wing aircraft built for ultra-long endurance

Most platforms are fitted with solar panels and supported by high-density batteries. At these heights, they can capture sunlight for extended periods, remain airborne for weeks or months, and function with little fuel or maintenance.

By cutting the distance between transmitter and user from hundreds of kilometres to just a few dozen, stratospheric platforms can provide fast, low‑latency links at much lower cost.

One platform can cover an area extending across tens or even hundreds of thousands of square kilometres. This makes HAPS particularly suited to sparsely populated places where laying fibre or building dense mobile networks would be uneconomic, including deserts, mountain ranges, remote islands and extensive rural areas.

Why satellites alone cannot complete the task

A satellite’s view from space covers an immense footprint. Although useful, this brings a severe compromise: operators must either support large numbers of users with limited bandwidth each or limit access to preserve acceptable speeds. Satellites also have to contend with atmospheric effects, more severe space weather and complicated routing requirements.

Low-Earth orbit networks such as Starlink reduce latency by flying closer to the planet than older geostationary satellites. Even so, they remain far above aircraft and are constantly moving relative to the ground, requiring connections to be transferred between satellites.

By comparison, stratospheric platforms occupy a narrow band of relatively stable air. They can hover, or at least circle within a confined pattern, over one area. On-board propulsion and sophisticated flight algorithms enable them to maintain position despite stratospheric winds.

An old concept gets a second life

The underlying idea is not recent. Telecommunications researchers began exploring high-altitude platforms during the 1990s. Test flights in the 2000s showed technical promise, although they were expensive. Alphabet’s Project Loon, launched in 2011, became the most prominent example, using fleets of balloons to transmit internet access to underserved areas.

Loon achieved several high-profile demonstrations, including emergency connectivity after natural disasters. However, the programme closed in 2021. The need to keep balloons over their required locations, manage strong winds, retrieve equipment and conduct repeated launches made the operating costs too high when compared with rapidly industrialising satellite constellations.

Three developments have since altered the equation: solar technology has advanced, batteries have become lighter and more capable, and telecommunications equipment has reduced dramatically in size. Together, these changes are reviving interest in the concept.

The new wave of stratospheric internet players

A number of companies now say they can achieve what Loon could not: remain positioned in the stratosphere for weeks while keeping costs commercially viable.

Company	Platform type	Altitude range	Notable capability
Sceye (US)	Solar helium airship	~20 km	Long-endurance, precise station-keeping
Aalto HAPS (Airbus, EU)	Solar drone (Zephyr)	Stratospheric	Record 67 days continuous flight
World Mobile (UK)	Hydrogen drone	High altitude	Bandwidth up to 200 Mbps

Sceye: a vast solar airship above the desert

US start-up Sceye has developed a helium airship measuring 65 metres in length and covered with solar panels. Intended to operate in the lower stratosphere, it carries telecommunications payloads and uses its own propulsion to remain almost stationary over a chosen location.

The company intends to prove that internet services can operate from the stratosphere, initially through trials in remote areas where ground infrastructure is limited or has been damaged.

Aalto’s Zephyr: powered by sunlight

Aalto HAPS, an Airbus subsidiary, has created Zephyr, a slim solar-powered drone with a wingspan of roughly 25 metres. Made from ultra-light materials, it flies above weather systems where turbulence is less severe and sunlight is more predictable.

Zephyr has remained in flight for 67 consecutive days, setting a record for an uncrewed aircraft. On missions of this kind, it can slowly circle above an area and effectively function as a mobile phone mast suspended in the sky.

World Mobile: challenging Starlink on price

UK company World Mobile is developing hydrogen-fuelled drones for high-altitude use, with a clear objective: reducing costs enough to make connectivity affordable for low-income communities.

Each platform is designed to offer around 200 megabits per second of bandwidth. To demonstrate the potential, the company makes a striking comparison. It estimates that nine such platforms could serve all of Scotland, covering around 5.5 million people, for approximately £0.80 per person each month.

By World Mobile’s estimate, high-altitude platforms could serve a whole country for less than one pound per user each month, vastly undercutting satellite subscriptions.

For comparison, a standard Starlink subscription in the UK is closer to £75 per month, excluding equipment costs. Performance would not be identical, but the difference illustrates how significantly the economics can change when infrastructure is 20 kilometres above users instead of in space.

Working with satellites and terrestrial networks

Stratospheric internet is not designed to displace satellites or land-based mobile networks. Its purpose is to fill the spaces between them.

• In high-density cities, fibre and 5G are likely to remain the quickest and most dependable choices.
• Across moderately populated areas, standard masts and microwave backhaul can handle much of the demand, while HAPS address weak-coverage locations.
• In remote locations, a small number of high-altitude platforms may be the only practical means of supplying broadband without major infrastructure expenditure.

The biggest challenge now extends beyond engineering. Regulators worldwide need to establish how HAPS will share radio spectrum with current services, coordinate with satellites, and comply with airspace and safety regulations. Without aligned rules, operators may encounter delays and fragmented markets.

Latency, bandwidth and other jargon explained simply

Three technical concepts are central to the discussion around stratospheric connectivity:

• Latency: The period required for data to travel from a device to a server and return. Low latency makes browsing feel more immediate, video calls smoother and online gaming more responsive. Since HAPS sit closer to Earth than satellites, they can offer latency more comparable with 4G or 5G networks.
• Bandwidth: The greatest volume of data that can travel through a connection each second. It is like the width of a motorway: additional lanes allow more cars through. One high-altitude platform can provide hundreds of megabits per second to be shared by users below.
• Throughput: The real-world speed experienced by users. It is affected by available bandwidth, the number of people sharing it and how efficiently the system manages traffic.

As HAPS cover limited geographical areas, operators can adjust capacity with greater precision than they can using distant satellites. This detailed control may be particularly valuable where demand changes with agricultural seasons, tourism or migration.

Risks, benefits and future scenarios

The expansion of stratospheric internet brings several risks. Long-lasting aircraft and airships create questions around airspace management. If a vehicle fails at altitude, its descent could present safety hazards in populated locations. Cybersecurity is another concern, since compromising one platform could interrupt service across a large area.

Conditions at 20 kilometres are calmer than those at commercial airline altitudes, but they are not completely stable. Platforms need to endure powerful winds, low temperatures and intense ultraviolet radiation for prolonged periods. Maintenance also requires difficult recovery and relaunch procedures.

Nevertheless, the advantages are attracting governments and private investors:

• Quicker emergency deployment after earthquakes, floods or conflicts
• Affordable connectivity for schools and clinics in isolated communities
• Backup connections when terrestrial infrastructure breaks down
• Assistance with environmental monitoring and border surveillance

One credible model would see countries combining infrastructure types: fibre in urban centres, 5G in suburbs, and stratospheric platforms for villages and farms that fall beyond the economic reach of masts and cables. Another approach would use HAPS for temporary “pop‑up coverage” at major events or in areas affected by long-term infrastructure damage.

For the moment, Starlink-style constellations still lead the discussion around global coverage. But as high-altitude platforms develop and regulatory frameworks become established, the prospect that the most effective internet connection may not come from space at all is beginning to resemble a business plan rather than science fiction.