Carrier-grade Wi-Fi: a backhaul bottleneck by 2019?

Real Wireless was recently commissioned by Bluwan to analyse the emerging trends and technological challenges that carrier-grade Wi-Fi is creating.

The study, Carrying Carrier Wi-Fi: The Technology Challenges, emphasized the increasing reliance operators will be placing on carrier Wi-Fi over the next five years. In fact, our analysis reveals that carrier-grade Wi-Fi hotspots will make up more than 80% of all available access points in 2019.

The study also identified that data demands being placed upon Wi-Fi hotspots will increase dramatically in the coming years, driven by data-intensive applications like streaming content. In fact, according to figures from Cisco, the world’s hotspots will need to support a staggering 63 exabytes of Wi-Fi IP traffic each month by 2018.

While there is little doubt carrier-grade Wi-Fi will form a crucial part next generation mobile networks, operators need to be aware of how important effective backhaul provision will become. In particular, the affordability of backhauling such dense network hotspots will become an increasingly serious barrier.

To avoid this, our analysis identified eight key technology criteria that Carrier-grade Wi-Fi needs to meet in order to ensure they are both cost-effective and operationally efficient to deploy. These key criteria are:

  • < $1,000 per link
  • Flexible planning
  • Flexible capacity allocation
  • Low Total Cost of Ownership per Mbps
  • > 500Mbps throughput
  • High scalability
  • Low latency: < 10ms
  • Quality of Experience 

In the report, we then benchmarked a range of wireless backhaul technologies against these criteria to assess their suitability for tackling this challenge.

For more information on the findings, our full analysis, and our conclusions, please download a copy of ‘Carrying Carrier Wi-Fi: The Technology Challenges’ here.

2G and 3G are dead, long live LTE

Earlier this month, Verizon CFO Fran Shammo finally confirmed that the long delayed launch of VoLTE on their networks will happen in Q4 of this year.

This signals a turning point in the technology; it’s been a long and slow road to get here, but we’re finally at the point where it is starting to infiltrate the mainstream conscious.

Both AT&T and Verizon have committed themselves to offering phones that can take advantage of the new VoLTE technology by Q4 – and I’d hazard a guess that means it is certain to be a standard feature in both Apple and Samsung’s latest generation phones. This in turn will undoubtedly mean their competitors are not far behind with their own offerings.

So far no real surprises. The more interesting question, though, is when will we see the first LTE only devices? After all, many operators and handset manufacturers have made no secret of their desire to turn off 2G or 3G networks.

For the operators, supporting these now legacy technologies not only occupies valuable spectrum, but adds additional infrastructure rollout and maintenance costs.

For handset manufacturers, the need to make use of 2G and 3G networks adds additional modem requirements and costs. These in turn negatively impact battery life and phone size. We’ve recently seen several new companies emerge offering “LTE Only”, thin modems at very aggressive prices, which no doubt has piqued the interest of manufacturers.

Obviously switching over entirely to LTE has only been made possible with the introduction of VoLTE. The lack of voice support has meant that a circuit switch fall back has been a requirement up until now, therefore 2G and 3G networks were a necessity.

Another key barrier up to now has been LTE coverage. Obviously, until this catches up, we’re unlikely to see any operator in a hurry to offer handsets that only support LTE, as this would severely impact their customers’ experiences.  But, as we saw in our recent work for the Scottish Government, the speed with which LTE has rolled out means it won’t be long until it catches up – our estimates put indoor 4G coverage in Scotland at 95% by the end of 2015.

Verizon originally forecast the introduction of LTE-only phones to their network by the end of 2014, a prediction that raised more than a few eyebrows. Their updated forecast now pushes this out to early 2016.

I think this is not only likely, but perhaps a necessity; should they wait any longer, the ecosystem will be in place for a competitor to take advantage of their delay.

Boris and 5G – will London really have it by 2020?

Boris Johnson has recently announced that London would lead the world in having 5G by 2020.

It’s not where you’d expect such an announcement to come from, at least not compared to the 5G stories so far. Boris certainly doesn’t run a cellular network, he isn’t responsible for telecoms policy or spectrum and (quite probably) won’t still be Mayor then.

But will London actually have 5G by 2020?

Cellular standards do follow an interestingly regular ten-year ‘tick.’ The first 1G analogue standards (e.g. AMPS,NMT) were launched around 1981; GSM first launched in 1991; UMTS 3G launched in between 2001-2003 and LTE was 2010.

So simplistically, you could say 2021 for commercial launch was about right, making pre-release demonstration on 2020 viable.

In LTE all the key requirements were agreed, and the basic technology outline had been accepted by 2004. However there is still a lot of mileage in LTE-A and the interim releases so 5G could be much later.

The other issue is the uncertainty over what exactly 5G is. With a wide set of use cases, a huge variety of architectures and technologies and, as yet, little consensus, it is hard to believe 5G will be a quick development.

This all makes Boris’ 2020 deadline look unlikely.

What I do think will happen is a demonstration and claims of “5G” in July 2020, driven by the Tokyo Olympics and the Japanese commitment to show something working. But even this may need to be taken with a pinch of salt, given the history of FOMA – a slightly incompatible pre-standard version of UMTS that launched in Japan in 2001, two years ahead of the ‘proper’ Release 99.

So while Boris could have his wish and demo 5G in London in 2020, it’s likely to be an engineering demonstrator with an instruction set in kanji shipped over from the Tokyo Olympics, and it won’t be the real 5G. The rest of us will have to wait at least another two years before we’re using 5G services – whatever they turn out to be.