Sunday, October 05, 2008

Future of Wireless - on the edge of chaos

I just attended a very elegantly presented talk by Linda Doyle at the Vodafone series on Mobile, at the Royal Academy of Engineering on the future of wireless communications especially with regard to spectrum use/allocation.

Summarising the range of design choices - there are three points on the spectrum (so to speak) Of design for allocation of this resource

1. command&control (top down allocation by central authority -think: just like IP address allocation/registries) (curent regime, mainly)

plus point: stability of "market" so planning for deployment of infrastructure (e.g. basestation/tower and backhaul, plus handset technology choice) Is long term assured, plus and strong isolation between different users

minus point: hard multiplex and slow rate of allocation/recovery cycle means frequently poor utilisation

2. spectrum trading/market

plus point: potentially better use and possibly leads to more innovation in use of spectrum, retains isolation between users

minus point: potential for instability and less long term certainty about ownership (financial markets today don't offer much hope with regard to stability as a feature of trading:) - hoarding, hedging, futures/dervtatives in spectrum could cause many problems

3. A commons

plus point: high utilisation, price efficiency (zero!)

minus points: tragedy (of the commons)
this, in my view, is overstates - the tragedy of the commons is a phrase from
anthropology referring to historical problems with common land being overgrazed:
too many cattle eat all the grass, grass dies, cattle die. but radio spectrum does't die when there's too much interference - just as with congestion in today's internet, people go away when the performance is poor, until it is "just good enough", and then make progress. maybe, if done right, a spectrum commons could be accessed dynamically per user like tcp.

B. Underpinning this was a range of very interesting new technologies in software radios which allow one to build a variety of flexible, cognitive devices which can measure (possible in a cooperative way) the current demand on spectrum locally, and using this, and possibly historical data, plan access for a device or set of devices. This then permits a more "fluid" approach than the current "rigid" allocations.

There were a great many questions, (along the lines of the plus/minus points I outline above). The slides were very helpful (using the Rubik's Cube model of spectrum over space, time) and contained a number of very well chosen artisitic metaphors to help with understanding this very complex design space.

I think that in the end, the idea of a mixed economy seemed popular (I would vote for 50% license exempt myself:). Some technologies like fiber radio, and diverse routing/antennae and cooperative relaying, may mean that re-allocations can share backhaul, thus allowing a dynamic allocation not to undermine long term planning of the infrastructure...

No comments:

Blog Archive

About Me

My photo
misery me, there is a floccipaucinihilipilification (*) of chronsynclastic infundibuli in these parts and I must therefore refer you to frank zappa instead, and go home