4G Thoughts

As part of the Parliament and Internet Conference, there was a panel discussion about 4G networks.

These are my notes on the session. I've applied the Chatham House Rule - mostly because I can't remember who said what, rather than any backroom skulduggery. Any errors are mine and mine alone.

Neither Vodafone nor EE have signed up to the Network Neutrality pledge. O2 have.

(Disclaimer, I work for Telefonica, these thoughts are my own.)
The Open Internet Code of Practice is the Government’s non-regulatory approach to net neutrality.

It basically says that ISPs will not discriminate between services, and if they do they can't call their product "Internet Access". It's all fairly sensible but O2 is the only mobile ISP who have signed up. This means the vast majority of 4G providers could start to block or degrade services which they think don't fit their business model.

Indeed, EE say they don't ban VoIP but want the right to filter such OTT services.

Ed Vaizey's form of "free market" capitalism states that because we have such good competition in the UK, people will be free to switch to other services which don't break the net. This conveniently ignores the fact that customers are tied in to 18 month long contracts (24 months if they want EE's 4G) and so are not able to easily switch.

Net Neutrality also offers some other interesting questions. VoIP and TVoIP have vastly different needs when it comes to latency - is it acceptable to prioritise a constant VoIP stream? Yes - assuming all streams of a similar nature are treated equally.

Of course, if those services are encrypted it could be very hard to tell which services are which.

Finally, emergency service calls are prioritised on the radio access and given network priority. Is that discrimination? Yes - but broadly acceptable in a net neutrality context.

Speed of Rollout

The 4G rollout is expected to be quick - much quicker than the 3G deployment. This is due to a number of factors.

• The auction has been brought forward.
• With EE having a head start, other operators will have to compete rapidly.
• Spectrum is expected to be much cheaper than the 3G bands. This means there will be more money for masts and networking kit.
• Mast sharing and relaxed planning permission should make setting up the network much more efficient.

Transco

There is only one electricity grid. One water network. One set of gas pipes. Yet different companies can sell different services at different prices across them. This idea is known as "Transco".

In the UK we effectively have only two mobile networks. EE (T-Mobile and Orange - who share with Three) is one, Vodafone/O2 is the other (they share masts, not spectrum and network kit).

All the major providers have outsourced the running of their network to NSN or Ericsson (an idiotic idea, but there we are).

Each company has to bid billions for new spectrum and spend millions on new masts and networking kit. Wouldn't it be more sensible if there were a national mobile network with O2, Vodafone, EE etc just acting as MVNOs? Each could buy wholesale access (and still run their own customer databases etc) at a vastly reduced cost and the public should benefit from improved coverage.

There seemed to be a lot of sympathy in the room for this idea. Apparently the Netherlands is investigating "National Roaming" which would allow your phone to roam on to a competitor network if you didn't have any signal.

One other point of interest - Ofcom apparantly want to see a non mobile operator in the 4G space. Perhaps someone selling wholesale or M2M access.

Speed

There are serious worries around real world speeds - as noted by EE claiming only 12Mbps downlink speed. Given the hype around the speeds is not likely to be met by real-world experience, this could deter customers.

The mobile industry may also be overselling capacity. ADSL suffers from insufficient backhaul - domestic ISPs bank on users only using the internet in short burst so don't buy enough connectivity. If we're all streaming movies all the time - speeds will suffer.

Price

Will customers pay?

There has been much derision over EE's pricing as being far too high and restrictive. Considering how long it took to build up a 3G customer base, can the industry attract customers to 4G when the coverage is patchy, phones are expensive, and battery life is worse? If they do - can they convince them to pay a premium for it?

Consider domestic ISPs. Pipex offered dial-up internet access for 50p a day back in 1996. Today, 16 years later, that £15 per month will get you "superfast" 24Mbps Internet access.

Technology brings prices down - even if they blip up temporarily. No doubt when other players enter the 4G market, prices will tumble.

Finally...

What will people do with 4G?
Most people expect "The same, but faster". Can we find new services, or are we stuck with better speed, lower latency?

That's not a bad thing to be stuck with, but I wonder what the "killer app" will be that makes people want a 4G phone in their pocket... We all thought that 3G's killer app would be video calling. How wrong we were!