Tuesday, 22 September 2009

Digital Britain: is it time for more radical spectrum policy?

Following the thoughts on the 'green-ness' of telecoms companies, it struck me that the forthcoming auction of digital dividend spectrum in the UK is a great opportunity for the Government and Ofcom to stop tinkering round the edges of both the green agenda and Digital Britain and take some positive action.

Auctioning the spectrum (suitable for LTE) to the highest bidder is a safe move, but it won't bring us any closer to either meeting the pledge to halve our carbon emissions by 2050, or the rather shorter term objective of giving everyone the option of connecting the Internet with reasonable bandwidth. Instead, I propose a scheme based on rewarding operators for being part of an industrial and social commons.

The carrot in this scheme is free spectrum for cellular LTE use. The value of this is yet to be established, however we have many data points to choose from, for example:
  • GSM licenses in the UK cost £142,560 per 2*200KHz slot per annum. This adds up to about £16m per operator;
  • 3G licenses were settled by upfront payments to HM Treasury of about £22b, split 5 ways (this was however, due to over excitement on the part of the bidding parties, amongst other things)

I prefer the first data point, or schemes such as that in India, where the license costs a small percentage of revenue, hence derisking investment (to a point). My suspicion is that Ofcom will favour an upfront capital payment for a chunk of spectrum. My prediction is that a 2*20MHz slot will cost the buyer between £300m and £500m.

In any case, in my scheme the spectrum is free, provided the following criteria are met:

  • The network must be demonstrably carbon neutral within 3 years of the license being awarded. This condition should include all operations and cannot be met by carbon trading or offset schemes;
  • All participating operators must agree to provide data on handset movement within cells to an aggregator, which will turn the information into a feed, similar to RDS, that enables efficient traffic routing applications. I see this as a similar 'social technology' to GPS;
  • Operators must commit to providing broadband to the final 10% identified in Digital Britain. Technical considerations will determine how fast this is, but I suggest a minimum of 512kbit/s, enabling eGovernment and other initiatives in the future. The range of LTE and the suitability of 800MHz should make this eminently feasible. I also suggest that a single rural network would be beneficial to achieve this (but that's another story)

These may or may not be the right criteria, but my opinion is that the sentiment is correct. We face an environmental crisis - that is nearly certain - and in the Government's eyes also a 'digital divide'. Conventional regulation and policy making will not jerk industry into action. It will not lead to innovation on green issues, or sudden decisions to act anti-commercially.

Neither do I believe that government should intervene in a command economy sense. What commercial entities need is a nudge in the right direction. Consumers will not do this alone as they are not experts in the technologies or supply chain for the services they love. It is up to the Government and Ofcom to set the UK on the right course and in this case, that does not mean taking hundreds of millions off the shareholders of MNOs.

Monday, 21 September 2009

Industrial commons & automotive social networking

I've been asked to contribute of a debate on 'green telecoms', specifically how the telecoms industry can enable other industries as they seek to reduce carbon emissions. A couple of thoughts I've had on the subject:

First, should the telecoms industry contribute to the industrial commons by making certain data available to all, in the same way that GPS signals are open to anyone with a receiver? Tom Tom's HD Traffic service is one service that makes use of data on handset movement to route traffic. Could this data benefit everyone, reducing jams and hence emissions?

Second, is there a benefit to a short range car-to-car communications system, a kind of automotive social networking that enables ECUs to communicate with each other? Besides the obvious benefits around collision avoidance, could such ad-hoc networking help manage traffic flow and even enable the sharing of services such as GPS routing and even in-car entertainment? Sounds interesting conceptually - the EU has even set aside frequency for it at 5.9GHz.

There's most definitely money in this for OEM manufacturers to put chips in every vehicle whether at build or as a retrofit. I also wonder whether there is a micro-payment model for exchanging access to services between vehicles. The challenge to this would be identity management, however accounts could easily be tied to vehicle registrations to get around this.

I'll feed back how the debate goes...