TIP 2018: what’s in it for Facebook?

At the Telecom Infra Project Summit 2018 we spoke to the Facebook execs behind the initiative to find out why they decided to get involved.

When Facebook first started talking about getting involved in in the telecoms industry via TIP and even developing novel wireless technologies such as Terragraph, it felt like a frustrated OTT going through the motions to light a fire under the sector. Facebook’s vested interest was clear: the better and more ubiquitous the connectivity, the more people will use Facebook.

As we explained earlier, a big part of this involves efforts to make telecoms infrastructure cheaper to buy, roll-out and maintain. In that respect TIP is a direct threat to the traditional big kit vendors, not only because tower networking costs probably equate to lower profits for them, but a major aim of TIP is to expand the whole telecoms ecosystem, thus creating additional competition for them.

In a couple of small media gatherings at the event we spoke to Jay Parikh, Head of Engineering and Infrastructure at Facebook, and VP of connectivity Yael Maguire. Parikh explained that TIP is not just a product of Facebook’s own connectivity needs but also of conversations he was having with operators two or three years ago in which they implored Facebook to get involved.

The biggest mutual problem faced by Facebook and the operator community is the exponential growth in traffic over networks combined with the increasing difficulty and cost of providing it. “We were worried that innovation was slowing down,” said Parikh, in reference to the collective concern felt at the time, one which the big kit vendors were failing to sufficiently address.

In response to persistent questioning about the return Facebook expects to get on its significant (but unspecified) investment, Parikh insisted that this isn’t a short term thing for his company. The strategic objective is to catalyse the telecoms industry and ROI will be gauged by the presence of novel connectivity innovation, as opposed to direct financial considerations.

It’s easy to be sceptical any time a company claims to be doing something for the greater good, but equally this would be a strange area for Facebook to diversify into if it was only looking for a new profit centre. Having said that the world’s dominant etailer now makes much of its profit from its cloud business so you never know.

Parikh kept his cards pretty close to his chest regarding any TIP financial metrics but it’s relatively easy to believe that a cash-rich Silicon Valley company might be prepared to spend money a bit more speculatively than a traditional outfit. Facebook considered its own fortunes to be intrinsically allied to those of the global telecoms industry, so helping it innovate is viewed as sufficiently self-interested by itself, for now at least.

When asked what the top priorities are for Facebook from TIP, Parikh cited the connectivity insights programme, which aims to give operators additional data to help operators make informed decisions derived, in part, from anonymised Facebook user data. Rural access work is also important as Facebook seeks its next billion users, and Telefónica’s work in Peru was cited as an example of this.

The third priority is Terragraph, which is positioned as an alternative to fixed wireless access delivered over unlicensed 60 GHz spectrum, of which there is plenty, with an emphasis on backhauling wifi. This is a key concern of Maguire’s, who noted that average video speeds are declining across the board thanks to the aforementioned imbalance between demand and supply.

Maguire explained that Terragraph started as a project designed to look into the viability of using the 60 GHz spectrum for backhaul. At such a high frequency there are a bunch of propagation challenges, with even oxygen itself contributing to signal degradation. But it turns out that if you get the precise line of sight alignment right and don’t try to transmit any further than 200m, then it can be used in much the same way we’re talking about FWA over mm wave for 5G.

In keeping with Facebook’s general tone on this stuff Maguire played down any direct antagonism between Terragraph and mm wave FWA, insisting they just wanted to offer up alternatives. I was also keen to stress that this technology is specifically intended for high bandwidth wireless backhaul. “It’s not a solves all problems technology,” he said.

So, in summary, Facebook says it’s not looking for any immediate return from its involvement and investment in TIP. Instead it expects to benefit from the telecoms industry innovating as a faster pace than it would have if Facebook hadn’t decided to get involved. Aside from justifiable scepticism about any company being so sanguine about immediate, demonstrable ROI there’s little reason not to take Facebook at face value on this, while also keeping a watchful eye out for mission creep as things progress.

Two tiers very evident in US telco rankings

OpenSignal has released some more granular insight on the performance of the top four telcos in the US and it is very clear there are two tiers; Verizon and T-Mobile US are best-in-breed, AT&T and Sprint are not.

When the firm released its State of Mobile Networks in the US report T-Mobile US was recognised as the leader for pretty much everything, but when you look deeper into the statistics in the individual regions, the competition is a lot closer than top-line figures would suggest. Verizon beats T-Mobile US in some areas, is level in others and not far off the pace in the rest.

While the leader in the race is a lot closer than perhaps it might have been initially presented, what is clear is that AT&T and Sprint are not in the same league. The tables below demonstrate this point quite effectively.

OpenSignal 4G

On the availability side of things, the distinction between first and second tier is very clear. What should be noted is that the last 12 months have seen a great level of improvement from Sprint and the team has said it will continue investments to improve this coverage. Perhaps it will close the gap over the next year, but that is not a promise it will be able to compete on the speeds front.

Interestingly enough, while the US is one of the best in the world for 4G coverage, it doesn’t compete on the global stage when it comes to download speeds. Again, we can see there are two tiers in the operator rankings when it comes to speed with T-Mobile establishing a bit more of a buffer over Verizon in the top tier, but the overall average falls below the 16.9 Mbps global average. Admittedly, many of the countries above the US are smaller and do not face the same challenges when it comes to geographical variety, but this is a pretty poor performance from a country which claims to be at the front of the digital revolution.

A couple of weeks back we noted the companies who are performing best in their individual markets are the ones which are investing in their networks and not getting distracted by other more colourful ventures. This is another story which adds credibility to the theory. With 5G just around the corner, we would hope AT&T doesn’t get drawn into a long and bitter legal battle with the Department of Justice over its acquisition of Time Warner as that could spell disaster for its standing in the 5G world.

Ofcom pats UK MNOs on the head for doing what they’re told

UK mobile operators vowed to hit Ofcom’s 90% geographical voice coverage target by the end of last year and they have. Good operators.

To commemorate this touching gesture of compliance Ofcom has written to each of them to say well done in a manner reminiscent of getting a commendation from a teacher. The specifics of the letters involve acknowledgment of their previous letter claiming to have done what they were told and confirming that Ofcom doesn’t think they were lying.

There are some intriguing variations within those letters, which detail the extent of UK geographical coverage by each MNO and even the date they sent their letter of claimed compliance. As you can see from the table below Vodafone, O2 and Three were all comfortably over the 90% threshold, and send their letters with plenty of time to spare. For EE, however, it seemed to go down to the wire and it just crept over the line at the 11th hour.

UK MNO geographical coverage

Inevitably Ofcom doesn’t want operators to rest on their laurels. There will be a 700 MHz spectrum band auction in 2019, we’re told, and Ofcom is proposing that spectrum should come with rural coverage strings attached. The precise nature of those strings wasn’t detailed but a balance needs to be struck between doing the right thing and turning that spectrum into a chalice, poisoned with onerous obligations.

Lastly Ofcom has published a document entitled ‘Enabling 5G in the UK’, which features a bunch of measures and top-tips designed to, well, enable 5G in the UK. 5G is the next generation of mobile technology, we’re informed, and Ofcom feels an obligation to muck in to the collective effort.

Apart from encouraging everyone with periodic pats on the head, Ofcom wants to make yet more spectrum available for 5G fun. On top of the aforementioned 700 MHz action and the stuff due to be bid on this year, Ofcom also wants to dig up some 3.6-3.8 GHz spectrum for auction in 2019. Apart from that 26 GHz seems to be the targeted millimetre wave band and there are some vague aspirations for 66-71 GHz too.

Ofcom 5G spectrum plans

There are some other initiatives, such as trying to make accessing base station sites easier, helping out with backhaul by stimulating fibre investment, and making sure European net neutrality regulations don’t get in the way of network slicing. On the whole this set of announcements paints a plausible picture of Ofcom being on top of things, which is presumably one of the main reasons for issuing it.

Mobile network experience in Scotland – in order to get it right we need to understand what is wrong

Telecoms.com periodically invites third parties to share their views on the industry’s most pressing issues. In this piece Brendan Gill, CEO of OpenSignal, looks at the complexities of accurately measuring mobile signals.

Scotland has been struggling to keep up with the rest of the UK in terms of digital connectivity for many years. Official figures from Ofcom show that a mere 17% of Scotland has 4G mobile coverage, compared to 60% in England. And although there have been several initiatives launched by the Scottish Government to address this – such as The Mobile Action Plan, or the R100 programme pledging to deliver superfast broadband access to 100% of premises in Scotland by 2021 – the gap remains visible in key connectivity metrics.

In most cases the debate has centred around how Scotland’s consumers are being left behind. But erratic connectivity, especially in mobile services is having an impact and hurting businesses as well.

It’s being measured wrong

“Fixing” Scotland’s connectivity issue can only start once there is a clear understanding of what the problem is. Taking it a step further, if problems are not identified and quantified correctly, they serve merely as a distraction, delaying the process of finding a solution. In the wireless world, the distractors are the measurements used to quantify 4G connectivity.

The way that coverage is currently measured has a lot of limitations. Talking about 99% population coverage paints a very rosy picture of a rather grim tale.  It’s vital that the right metrics are in place to truly assess – and address – the problem. Only then can the discussions begin on how it can be fixed.

So how does the industry get it right?  One approach, from the likes of OpenSignal, is to look at time. Specifically, the percentage of time users are able to connect to an LTE signal. This draws a much more realistic picture of the everyday mobile network experience by collecting and analysing data from smartphones wherever their owners are: indoors or out, in the city or the countryside, day or night. The metric, called 4G availability, shows  Scotland (as well as the UK in general) still has a long way to go before reaching that coveted 99%.

By relying on real-world data, the availability approach eliminates the pitfalls that other coverage metrics so easily fall into: such as disregarding population density (i.e. geographic coverage) or failing to measure coverage indoors or at any location other than your home (i.e. population coverage).

Indoor coverage needs to be part of the equation

In fact, the majority of service fluctuations and black spots occur inside homes and offices. And many operators currently lack the means to track, let alone address the issue.

Population coverage is often talked about, and typically looks at: is there a signal at your doorstep? That would be great if all mobile devices were only used on doorsteps. But the reality is mobile users go inside, and there is a huge differential between the signal received outside and inside homes.

Searching for a signal at home, in the supermarket or in the Palace of Westminster itself should not be a source of frustration.  But as we look at the coverage map below (built based on real-world user data), it’s clear that Parliament’s indoor coverage is close to non-existent, while the surrounding areas all indicate strong signals.

westminster coverage

2 days less connectivity

The reality is that national 4G availability statistics are troubling – UK-wide we are talking about between 58% and 78% – and that is between the different providers. There is a big difference between this and the 99% population coverage figure that so many in the industry like to use.

As for Scotland, we were seeing around 7% less time connected to 4G compared with the UK nationwide average. It might not seem like a lot, yet 7% of the time is potentially two days per month when a mobile user in Scotland is not connected to 4G, but the UK-wide average is.

There is clearly a long list of unique challenges hindering 4G connectivity in Scotland, ranging from population spread to difficult terrain; but before the industry can start working on solving the issue, they first need to make sure that the metrics are right.

 

Brendan Gill_OpenSignal CEOBrendan Gill is the CEO of OpenSignal, a company he co-founded in 2010. He has spent over 10 years providing solutions to help people understand and improve mobile service and experience. Prior to OpenSignal, Brendan was part of the team that launched RepeaterStore in 2007, which provides signal boosting solutions to improve wireless cell and data reception in buildings, homes and vehicles. Brendan is listed in the Global Telecoms Business Power100 2017 as one of the most powerful names behind the telecoms sector. He is an accomplished speaker and has presented at leading industry events including: CTIA, Mobile World Congress, TechCrunch Mobile and the Qualcomm CEO Summit.

Passionate about empowering and supporting entrepreneurship, Brendan founded BetaFoundry, an accelerator programme offering mentors and advisors for students to encourage them to choose an alternative to the standard career path. In addition, he is one of the TechStars London mentors, and has taken the FoundersPledge to encourage tech entrepreneurs to donate to charitable causes. Brendan holds a degree in Physics and Philosophy from the University of Oxford.

Speeds aren’t getting faster, but 4G is coverage is expanding – OpenSignal

OpenSignal’s most recent State of LTE report claims that while download speeds across the world have started to plateau there is promising work being done to improve coverage.

While this will frustrate urbanites wanting their cat video fix, farmers wanting to surf for images of milk will be thrilled.

The last couple of years have seen global 4G download speeds improve quite considerably, though some nations plodded along (the UK being one of those), but this trend has been slowing recently. While we might point the finger at lazy telcos who might be more interested in collecting profits than improving customer experience, they are generally improving geographical coverage. OpenSignal has said three quarters of the European countries included in the research can provide LTE connections to mobile users in more than 7 out of 10 attempts. Only Armenia, Belarus and Russia had average availability scores below 60%.

Perhaps even more encouraging is the fact the Western European nations have finally pulled their finger out and are now offering better availability, though users in France, Germany, Ireland and the UK don’t average download speeds above 25 Mbps. One possible explanation is the focus on 3G in these countries in years gone. Early and powerful deployment of 3G satisfied the data demands of the population for a significant amount of time meaning the consumer demand for 4G was delayed. This bit the telcos recently, though this negative position does seem to be turning.

Looking specifically at the UK, in November 2016 its availability score was 57.9% though this has increased to 77.3% in the most recent report. There have been a number of projects in the last 18 months to encourage investment in the not-spots, but this is an encouraging statistic to prove it is simply more than telcos and MPs preaching for PR points. Recently the government announced an initiative to use churches to house equipment to improve coverage and such moves will only keep these statistics moving north. There is still work to match the best in the world, and Europe (Norway’s coverage exceeds 90%) but the trends are positive.

It should be worth noting that the slow-down in speed increases is a global trend, but when you look at the European nations they are closing the gap with the leaders out in Asia. OpenSignal believes that if the Asian markets continue to stall while Europe continues to steadily grow, the distortion could be addressed before too long.

While reports such as this are important to measure whether the telcos are actually doing anything notable, we decided to run our own mini survey. Your correspondent asked several friends to conduct a short speed test in a variety of locations, and the results are quite interesting. The tests were taken at 3pm (Tuesday 20 January) from a mix of different environments including working from home, on the road, central London and the Welsh countryside.

New Ofcom mobile coverage report yields fresh political posturing

The UK telecoms regulator has moved the goalposts for measuring geographical coverage, in turn giving politicians the chance to look like they’re doing something.

In its Connected Nations Report 2017, Ofcom announced it has been doing some of its own testing and found that the state of UK geographical coverage is worse than it had previously thought. This is significant because geographical coverage has been a condition of some licenses and it has also been a focus for politicians ever since David Cameron got the hump about not being able to check the football scores on holiday.

“People have never relied so much on their phones in daily life. As a nation, we are using 13 times more mobile data than just five years ago,” said Ofcom CTO Steve Unger. “While the industry works to improve mobile coverage, it’s vital people can get a trustworthy picture of reception across the UK. Using our tools, mobile users can see which network offers the best service in areas where they live, work and travel, before they take out a new phone contract.”

Ofcom state of mobile

Speaking to operator sources Telecoms.com learned there are many question marks around this shifting of the goalposts by Ofcom and the subsequent political opportunism by Lord Adonis, the Chair of the National Infrastructure Commission – a body that seems to have been set up to moan about UK telecoms.

Firstly, Ofcom has been far from transparent about its methodology. Its new criteria for ‘coverage’ are “Nearly all 90 second voice calls to be completed without interruption and speeds for nearly all data connections to be fast enough for users to browse the internet and watch mobile video effectively.” The latter, apparently, requires a minimum data throughput of 2 Mbps. But how sure can Ofcom be of the extent of those areas that fall short? Did it measure every square foot of the country and if so how?

And then there’s the unilateral nature of the revision. When it imposed geographical coverage obligations on operators originally there would have been an agreed set of criteria, which those operators have presumably been striving to fulfil. It hardly seems fair to both shift the goalposts and then immediately condemn operators for flouting the new criteria.

Which brings us onto Adonis. He wasted no time whatsoever in writing an open letter to Sharon White – the head of Ofcom – lamenting how terrible it all is and, of course, how much he cares about the poor, old UK population that has to struggle under the yoke of this medieval infrastructure.

“Despite licence obligations that were intended to provide coverage to 90% of the UK’s landmass by the end of 2017, large parts of the country remain without reliable coverage – with almost a third of the UK’s geography unable to receive a signal from all four operators,” wrote Adonis.

“A range of policies should be considered including, but not limited to, re-examining the case for roaming in areas where there are ‘not spots’, making better use of existing spectrum and encouraging MNOs to share masts where possible.”

The first quote reveals another major problem with this Ofcom manoeuvre – the need for 90% of the country to be covered by all four operators. We would imagine that Three, which is seldom slow to moan, is likely to be especially aggrieved at this since it has the smallest network. And, for that matter, why is it so important that we’re able to choose between four providers when we’re at the top of a Welsh mountain?

Then Adonis starts banging on about making better use of existing spectrum. We have to wonder what existing spectrum he has in mind and how he thinks that spectrum is being under-used. He presumably knows that higher frequency spectrum won’t address coverage issues so he must think operators are sitting on sub-1 GHz spectrum, which seems unlikely.

Ultimately, when the state starts grandstanding about telecoms provision you have to question how it positions it. If it’s a utility then surely the government should be doing more to help resolve these issues than just moaning, such as facilitating sites on public land. If it’s a private enterprise then it should either chuck it some public money or shut up. Sadly neither seem likely.

EE wants us to focus on ‘time on 4G’ as the main network metric

UK operator EE has a rich history of trying to influence the way mobile networks are judged by the state and the public and this is its latest effort.

It’s actually part of a campaign EE CEO Marc Allera has been waging for over a year, kicked-off with an unsolicited Christmas letter he sent to his competitors. In it he announced his ‘clear on coverage’ initiative that urged the industry to shift from talking about population coverage – which is already close to 100% – to geographic coverage.

Now Allera wants to shift the goalposts once more, with ‘time on 4G’ – i.e. the proportion of the time a user spends connected via 4G, as opposed to older generations – as the definitive measure of how good a network is. Presumably Allera is confident that this metric will flatter EE.

“Our industry has to get better at giving customers the information they need to make an informed choice about the mobile network that best suits their needs,” said Allera. “I banned misleading population coverage measurements at EE a year ago, and we’ve seen real change since then. The fundamentally misleading claim of ‘99% coverage’ is very hard to find on mobile operators’ websites today.

“We are introducing new measurements today that will give us a more accurate view of our customers’ network experience than ever before. And we’re working with Ofcom to improve the quality of information that’s available to customers, as we share their ambition to improve transparency around network performance.”

The EE ‘time on 4G’ metric for an EE punter with 4G calling (VoLTE) enabled on their phone in an urban area is 96%, we’re told, falling as low as 79% in some rural areas and to 70% if 4G Calling isn’t enabled. 4G can be pretty moody on the train too, as this correspondent can confirm.

EE claims first 4G ‘air mast’ for ad hoc remote connectivity

A 4G base station hanging from a balloon was used to enable mountain bikers in Wales to do a 360-degree live stream of their antics.

Say what you like about EE, but the UK operator knows how to put on a publicity stunt. Whether its wallowing in mud at Glastonbury or skiing in Scotland, EE wants everyone to know how good it is at covering remote UK locations and will go to great lengths to prove it.

It seems the only thing missing from the Red Bull Foxhunt downhill mountain biking event in Snowdonia was the ability for participants to livestream their attempts to seriously injure themselves via 360-degree cameras. Enter EE and its balloons.

The recent event apparently saw the first commercial use of a balloon-based ‘air mast’ technology called Helikite, even though it’s neither a helicopter nor a kite. It provided 4G connectivity for EE subscribers and a wifi hotspot for everyone else. To show it delivered the promised connectivity EE got a rider called Juliet Elliot, who is apparently an Instagram star, to hand over her recording, which you can see below.

“We have 4G coverage in more places than any other operator and are going to extraordinary lengths to connect communities across the UK, but when we saw the remote location of Red Bull Foxhunt we knew we had to go even further – and our ‘air mast’ technology was the perfect solution to provide coverage on demand, keeping everyone at the event connected,” said EE CEO Marc Allera, improbably all in one breath. “This is the first time anywhere in the world that a Helikite ‘air mast’ has been used to provide complete 4G connectivity to consumers, and it’s a sign of how far we’re going to keep everyone connected.”

Outside of publicity stunts it remains to be seen how important this Helikite thing will be. Balloons clearly have a role to play in emergency connectivity, such as recently in Puerto Rico, but for permanent coverage surely the good old land mast remains the best bet.