Maybe 5G isn’t better: Computer Weekly Downtime Upload podcast

Arthur C Clarke’s profound paper from 1945 presents the concept of geospatial orbit, making the claim that with three satellites at such an orbit around the equator it would be possible to achieve global communications.

Today, the cost of launching satellites has fallen dramatically, to the extent that it is now feasible to operate a constellation of low Earth orbit (LEO) satellites, capable of delivering broadband to anyone and everyone.

But is this technology overkill, especially as 5G mobile networking is well on the way to establishing itself as the next generation of mobile communications?

The 5G half-truth

Bill Ray, a vice-president analyst at Gartner, believes there is a big problem with 5G. Ray is the lead author of one of Gartner’s so-called “maverick” research notes, offering a different perspective to the norm regarding the adoption of 5G mobile networking.

Gartner has forecast that by 2027, 35% of large enterprises will gain the majority of their new customers from areas that lack reliable connectivity. Ray’s research suggests that 5G cannot deliver on its application promises outside small islands of commercial deployment and vertically integrated examples.

In this podcast, Ray speaks to Computer Weekly about the challenges of the 5G roll-out and why he believes it is not delivering the connectivity required by businesses and consumers. “5G is a technology looking for a problem to solve,” he says.

According to Ray, 5G was designed by mobile network operators that wanted a reason to expand their operations and to increase their revenue. “It wasn’t designed to solve a specific problem. Nobody sat down and said 4G is not good enough and therefore we need to make 5G,” he says.

“The overriding issue with 5G is it is not better than 4G – it does not change the experience, it doesn’t make 4G better and it doesn’t improve internet access”

With the industry operating on a 10-year cycle, “we’ll have 6G in 2030”, Ray adds. “We still don’t know what it will be used for, but we will have it.”

While the headline-grabbing figure for 5G connectivity is the 20Gbps peak bandwidth number, Ray says that even the sustained bandwidth figure of 100Mbps offered over 5G exceeds the bandwidth that anyone can consume on a mobile phone.

“I think the overriding issue with 5G is it is not better than 4G – it does not change the experience, it doesn’t make 4G better and it doesn’t improve internet access,” he says. “Even if you’re streaming all the time, there’s a limit to how much you can stream. Once you get beyond about 70Mbps or 80Mbps, there’s nothing more you can do.”

While 5G makes it technically possible to download gigabits of mobile data per second, Ray’s view is that people are unable to consume this much data. “In fact,” he says, “you can barely put this amount of data into the phone’s memory at that speed, so you’ve got all this data rushing in, but you can’t do anything with it.”

While there are applications that do need extremely high bandwidth, today’s smartphones require far less. In fact, research from Gartner suggests that most enterprise applications, including streaming applications and video-conferencing, only require about 6Mbps, which falls well within 4G territory.

Despite the high-bandwidth headline figures touted by the mobile operators, Ray believes the main benefit of 5G networking is its ability to provide ultra-reliable low-latency communications (URLLC). “Rather than being connected to a base station, your mobile phone, car or drone connects to three or four base stations simultaneously. That way, if a base station disappears or the connection gets broken, your connection remains intact. It just slows down. This way, you can have completely reliable connections,” he says.

However, while it is in the technical specification for 5G, Ray says URLLC has not been deployed and we are still years away from having it.

The satellite era

In the research paper, Ray and fellow Gartner analysts Roger Williams, Alfonso Velosa and Marty Resnick discuss why satellite internet has an opportunity to provide ubiquitous connectivity in a way that puts 5G to shame.

The ability to use low Earth orbit has significantly reduced the cost of deploying satellites in space. Ray says these satellites are also now much cheaper.

“Ironically, the reason they’re cheaper is because of mobile phones,” he says. “We spent the last 20 years making antennas, sorting out radio frequencies, developing processors and batteries and all the stuff that goes into a smartphone. This technology is perfect for satellites: it is miniaturised, lightweight, robust and waterproof.”

LEO satellites today are mainly launched using rockets. To cover the whole planet, a satellite operator needs hundreds, possibly thousands, of satellites. Fortunately, rockets have improved significantly, making deployment easier. Virgin Galactic has even tested launching a missile with a satellite as its payload, attached to a Boeing 747 aircraft.

The Falcon 9 rocket from SpaceX, the company founded by Elon Musk, recently launched 50 LEO satellites in one go. Starship, the company’s latest rocket, is currently being tested. When it is ready, Ray says Starship should be able to take several hundred satellites into low Earth orbit on each launch.

As the number of satellites that can be deployed per rocket launch increases, the cost goes down. According to Ray, SpaceX’s Musk has ambitions to bring the cost of launching a satellite down to a very optimistic figure of $10 per kilogram. So, while hundreds of LEO satellites may be needed, the satellites being manufactured by SpaceX weigh just 260kg, which means that it may eventually cost just $2,600 to launch them.

Why use satellite internet?

Ray says Starlink, the part of SpaceX specialising in satellite internet, is already offering home internet access to half a million people anywhere in the world. “You can get a Starlink dish, put it outside and get internet access between 100Mbps and 200Mbps for $110 a month. This is already happening and it is changing people’s lives today.”

SpaceX is not alone. China has SatNet, a constellation of 12,992 satellites. The UK’s OneWeb 648-satellite constellation has now merged with EutelSat, which operates a 36-strong fleet of geostationary orbit (GEO) satellites. The European Union has announced €2.4bn funding as part of a €6bn European satellite constellation. And through Amazon’s Project Kuiper, the e-commerce giant is planning to launch 3,236 LEO satellites using rockets from Arianespace, Blue Origin and United Launch Alliance (ULA).

“Suddenly,” says Ray, “it’s starting to make sense to have these satellites in low Earth orbit.” 

Another significant opportunity for satellite internet is mobile backhaul. For instance, Ray says KDDI in Japan is deploying 1,200 base stations with satellite backhaul via Starlink. There is no need for high-speed fibre connectivity – all that is needed is a satellite dish and power. The electricity can even be provided via a diesel generator or solar panels, enabling operators to deploy mobile internet base stations almost anywhere.

“It now becomes much cheaper to roll out mobile phone infrastructure,” says Ray. While in the past, satellite communications would have been the only option in remote locations, now, anywhere that requires fibre connectivity to a base station to be installed can avoid this unnecessary cost by using satellite connectivity instead.

Direct mobile phone connectivity

“You can go one stage further, because a mobile phone can send a signal about 500km,” says Ray, which makes it possible to use mobile phones to connect directly to a satellite without the need for any mobile infrastructure.

In fact, that is exactly what satellite mobile operator startup Lynk Mobile has done. In April 2022, Lynk launched and deployed Lynk Tower 1. Charles Miller, CEO and co-founder of Lynk, says the company is positioned to begin commercial service later this year and provides what he claims is “the world’s first true satellite-direct-to-phone service to citizens across the globe”.

Nokia recently signed a deal with another startup, AST SpaceMobile, where its AirScale Single RAN equipment will be used with  AST SpaceMobile to provide mobile services to new and existing subscribers in regions currently not served by terrestrial communications networks. This includes connecting devices globally on land, at sea, or in flight, says AST SpaceMobile.

Discussing the partnership in July, Scott Wisniewski, chief strategy officer at AST SpaceMobile, said: “In the coming months, we are scheduled to launch our BlueWalker 3 test satellite into low Earth orbit, which has a 64m² phased array antenna designed for direct-to-cell connectivity. With this satellite, we plan to conduct testing all over the world with leading mobile network operators, leveraging Nokia’s technology solutions on the ground.”

Challenges facing satellite broadband

The big question, says Ray, is the economic viability of satellite broadband.

“Starlink plans to deploy more than 30,000 satellites. With 30,000 satellites, each having a lifespan of five years, just to maintain the network you’re replacing 500 satellites a month,” he says.

Given that the current Falcon 9 rocket is capable of launching 50 satellites each time, Ray says Starlink would need to launch satellites every three days just to maintain the network. Of course, there will be larger launch vehicles, such as SpaceX’s Starship, if this becomes operational. But these are unknowns.

There is also the prospect of manufacturing LEO satellites in space, but this is in the realms of science fiction. “We are pushing the boundaries,” Ray adds. “We are talking about manufacturing in space and flexible configurations. We need to do these things because the opportunities are so vast.”