The idea of using high altitude platform systems (HAPS) has been discussed for at least 30 years. The benefits of coverage from the skies are obvious – much larger cells and much greater probability of line-of-sight propagation. The challenges have been how to economically keep platforms airborne. Ideas have come and gone. Google’s Loon project tried using balloons over Africa but found that keeping them in the right location was very hard. Airbus’s Zepher project (now Aalto Zepher) is close to commercial operation after many years of development and trials. And now there are many other concepts being trialled from tethered aerostats to powered balloons and hydrogen-powered aircraft.
The reasons why so many companies continue to develop HAPS solutions are two-fold. Firstly, the need is still very clear. Coverage of rural areas in many countries remains poor and has mostly stalled as operators aim to reduce their capex. Secondly, new technologies are emerging that allow HAPS platforms to be deployed at lower cost. These include ways of keeping aerostats airborne for longer, pilotless drones and aircraft and new antennas that allow HAPS to project large numbers of cells onto the ground.
There is a strong need for HAPS and promising new technologies, but the path to profitable operation is still unclear.
Direct to device satellite muddies the waters
In 2022 Apple launched the iPhone satellite emergency messaging solution, demonstrating that direct-to-device (D2D) communications from a satellite was possible, albeit only for messaging applications. Since then, many companies have promised capable D2D solutions including SpaceX and AST.
If D2D could provide an acceptable alternative in areas where cellular coverage is lacking then there would be no need for HAPS. But questions remains over whether D2D can deliver more than a “2G” level of service and whether it has sufficient capacity for all of those who are outside of cellular coverage.
It is very hard to determine the capacity of a D2D solution without knowing how much spectrum it has available. Apple’s solution is highly constrained as its partner, Globalstar, only has a small spectrum allocation. SpaceX and AST plan to use cellular allocations rather than satellite spectrum, but mobile operators typically do not have much spare spectrum that they can dedicate to D2D solutions.
Using most-likely assumptions, D2D remains a 2G-level service for deeply rural areas. There is great value in this, but it leaves a gap between terrestrial coverage and LEO coverage where there is a need for a 5G-level solution with high capacity. This could be where HAPS fits. But until the parameters of a D2D service become clearer, there is unhelpful uncertainty around the extent and value of HAPS.
While technologies advance, commercial models do not
Mobile operators have not, so far, shown much inclination to embrace HAPS. While some, like Vodafone, have invested in D2D operators, HAPS investment has come from VCs and entities like Softbank.
At first glance this seems odd. HAPS potentially provide a lower-cost way to expand coverage, and indeed, if a number of existing terrestrial rural cells could be replaced by a single HAPS station then there could be a cost reduction. Why would an operator not want to simultaneously expand coverage and reduce costs?
In practice, most operators have concluded that coverage expansion brings them little economic benefit once most subscribers have coverage for most of the time. After that, subscribers tend to select operators on the basis of cost and other special offers, often assuming that all operators have broadly similar coverage.
Lower cost is something operators are keen on, but many are tied into long-term contracts with TowerCos or are required to take a complete portfolio of cell sites and cannot “pick and choose”. Further, most HAPS solutions are not yet commercially available and may require development funding. Operators typically do not expect to fund development, having grown used to buying fully developed solutions from the major equipment manufacturers. So while HAPS may be lower cost in principle, they may not be in practice.
Even if operators were keen to embrace HAPS, the models for delivering rural coverage are changing from competition to collaboration. Rural coverage is now often subsidised by governments, either directly or indirectly, and in order to get the greatest returns for their funding they are asking for shared networks, at least involving passive sharing of masts and in some cases active sharing of infrastructure as well. This means that the operators need to agree and collaborate on new towers. While they could agree and collaborate on HAPS solutions instead, the variety of different platforms and the uncertainty over which approaches will be best, make collaboration problematic. Simpler to all agree on doing more of what they have done in the past. And some governments make matters worse by having targets for numbers of masts deployed rather than area covered.
There are alternatives to operators owning and deploying rural coverage solutions. Deploying HAPS is a form of mast deployment, and many masts are now deployed by TowerCos. A TowerCo could, instead, deploy a network of HAPS. But the incentives for a TowerCo are limited. The more masts they deploy the more rental they can get. Replacing masts with HAPS would be a difficult business case for many unless they can sell “square kilometres covered” rather than mast sites rented.
In response to these issues, some HAPS providers have also become network operators. They both build the HAPS platform and then fly it, offering services. But they typically cannot offer services direct to consumers since they will not own the spectrum, and in any case setting up a consumer offering is expensive, requiring marketing, customer acquisition and care and more. Instead, they offer services to operators. But this is has all the same issues as the operators deploying themselves, and if the service is offered to all operators then none gain a competitive advantage.
LEO integration?
Apple has demonstrated a completely different coverage model with its iPhone-satellite emergency messaging. It is providing rural coverage to iPhone users as part of the overall value package of owning an iPhone. It can do this because (1) the service does not require access to mobile operator spectrum, instead using satellite spectrum assigned to Globalstar and (2) it can build into the handset the “rules” for using this service, which are broadly to offer satellite connectivity when the phone has no other forms of connection.
Apple, or indeed any other company, could do the same with HAPS. They could deploy their own network of HAPS devices, perhaps flying drones across large areas, and programme their devices to access the network. Indeed, Google owned and operated the network of Loon balloons, not a mobile operator. Issues of spectrum arise, but there may be solutions such as using satellite spectrum on the basis that satellite coverage will not be needed where there is HAPS coverage or using unlicensed spectrum.
But the iPhone satellite service does not appear to have provided a significant advantage to Apple – the fact that no other handset manufacturers have copied it suggests that it is not seen as a “must have” feature. And it is much harder for a non-handset manufacturer to make the service work because they may not be able to get the frequency band used integrated into the device.
Those LEO operators working with mobile operators seem to believe that they can deliver a suitable service directly from their satellites and are unlikely to want to take on a HAPS network themselves.
The way ahead is an enlightened government
As noted, most rural network expansion is now partly funded by government. A government ought to want to deliver the greatest possible coverage for their chosen level of investment, and HAPS is likely to deliver much greater coverage than terrestrial cells.
An enlightened government would be prepared to fund HAPS development and deployment as long as the total costs were less than a terrestrial cell deployment. But this is a quite different approach of giving a consortium of mobile operators a sum of money and telling them to get on with it. Governments dislike picking winners and struggle to invest in companies based in other countries. Most HAPS developers appear to currently be located in the UK, US or Europe. So, for example, African countries, which often have great need for more rural coverage, might well be unwilling to invest in companies based in the US. Of course, there are ways around this – a sufficiently profitable guaranteed contract for sales could be used by a HAPS company to raise the funding needed to complete the development.
The best approach of all might be for all those governments who are sufficiently concerned about coverage to want to improve it to get together and collectively fund the development of multiple HAPS solutions, perhaps on some sort of competitive basis, with the promise that they would then part-fund the deployment of coverage across their rural areas. The collective nature of this endeavour would overcome the nationalistic issues of promoting manufacturers and ensure the cost per country was small – likely only a few $10’s millions.
Then each government could run a national competition, asking for tenders for delivering its required rural coverage. This would incentivise the use of the lowest cost technology without the government having to select that technology itself. It would allow operators, TowerCos and others to bid, overcoming issues identified above. Government could even enable shared use of operator spectrum (since these are areas where the operators are not providing coverage) allowing non-operators a chance to deploy and operate service.
Might this happen? Collective government action seems unlikely unless some third party steps in to convene. This could be a body like the ITU, the GSMA or even a charitable entity with objectives to deliver better coverage such as the Marconi Society. Perhaps if the HAPS developers could organise themselves collectively to appeal for such an approach it might stimulate bodies such as these to offer it.
Early flight threatening to stall
It seems illogical. HAPS promises to both deliver better coverage and lower operator costs. Consumers are very keen on both as are governments. And yet a route to deliver these benefits is hard to envisage.
This issue is broadly one of market failures in a world where the structure of the mobile marketplace is no longer ideal for the utility-like entities that operators are increasingly becoming. Market failures imply the need for government intervention but it is hard to convince governments and then to get them to act in an optimal manner, especially where there are nationalist issues.
HAPS developers are trying hard to find a way through and may succeed by finding entities that are willing to fund them. That would be welcome although it would take longer, and fewer solutions will emerge than would otherwise happen.
There has to be a better way.