Broadband wireless goes to the max

Rupert Goodwins: A new wireless standard could do for connectivity what the PC did for computing, but we'll have to move fast

If you'd asked anyone ten years ago what would improve their computing experience, the answer would most likely have been a faster processor, bigger hard disk or more memory. Oh, and nicer graphics with better sound, if it's not too much trouble. These days, we've got all that. There are still respectable numbers of people strapping go-faster bits to their computers, but it's more in the spirit of custom car nuts than fundamental advances in usability. Now, everyone's dearest wish is for faster Internet connectivity.

The trouble has been that while anyone can make and sell hardware and software, it's been much harder to set up ways to access the Internet. BT has been most efficient in its inefficiency, maintaining a de facto monopoly on ADSL and ignoring any customer in an inconvenient location. Cable is for urbanites only. It's as if the PC market in the early 90s was limited to IBM and Apple, neither of whom could be bothered to sell you a computer if you lived more than a couple of miles from your nearest post office. So we've been stuck with expensive, limited and rather frustrating access while the rest of the IT world moves on at speed.

Of late, there's been a hint of relief. Third-generation mobile phone systems were going to provide universal broadband access -- shame nobody told the mobile phone companies. Wi-Fi and related ideas have got lots of people setting up rural wireless broadband companies and scrambling up masts, bypassing the BT bucolic blockade of the last mile but still stuck paying huge leased line bills. What we need is some magic device that does for connectivity what the 8086 processor in the PC architecture did for computing: a basic building block that does the job and is open to all.

The selfsame thought occurred to Intel earlier this year when it was one of several companies that revitalised the long forgotten Wimax consortium. Wimax -- Worldwide Interoperability for Microwave Access, don't you know -- is there to promote development of 802.16, yet another set of interrelated radio networking standards. But stay awake at the back: 802.16 is to 802.11 what the M25 motorway is to the Basingstoke one way system. Designed to operate over a number of bands from 2GHz to 66GHz, 802.16 can work over 30 miles and pump data at speeds of up to 70Mbps.

Originally intended for 'last mile' fixed wireless broadband links, the standard has already grown mesh additions to let stations relay data for others, and work is underway for mobile working. This week, Intel announced it would be designing 802.16 silicon: it's not alone, and other standards espoused by the chip giant haven't always taken off, but as a statement of intent it's very significant. If 802.16 takes off, it'll change the face of broadband.

The idea is that it'll become so mass-market that component costs will fall to the point where individuals can afford base stations -- each of which will be able not only to receive network access, but forward it as part of a mesh. When the mobile parts of the standard -- 802.16e -- are finalised, the Wi-Fi versus 3G argument could seem as antiquated as VHS versus Betamax. As for the expensive leased line into the backbone: if you can afford a 70Mbps link to your friend thirty miles down the road -- a connection with almost no running costs -- then the telcos can take their leased lines and turn them into skipping ropes. ADSL pushed the boundary of expensive megabit connectivity back towards the backbone: 802.16 can do the same for the next stage.

There is one potential downside. At the moment, the rural broadband movement is vibrant but vulnerable -- the companies forming to provide the service are typically small, enthusiastic and by no means overfunded. This should be precisely the sort of organic, heterogeneous environment in which a thousand flowers bloom and from which the best solution evolves: we're still benefiting from the times in the 80s when every month brought a new, incompatible but potentially exciting home computer design. When the PC standard took over, the experience was there to take a common hardware and software base and create an entire industry independent of the old guard. 802.16 could provide the same fertile ground.

If the big telcos wake up and smell the photons, they could still stymie the small guys. An aggressive, enthusiastic, innovative push into the areas they've ignored, based around a mix of open standards and carefully contrived lock-in additions, could change the equation. They could realise what the end game will look like and leap ahead, but they'll have to be fast.

Fortunately, words like innovation, enthusiasm and speed are rarely to be found in the same boardroom as telco executives. This corporate conservatism, coupled with the many failed experiments in fixed wireless access that litter the big boys' attics, means that there's still a window of opportunity for a combination of 802.16 and good thinking by the rest of us. We could still see a revolution in connectivity that drives a burst of inventive, exciting change in real, usable technology, much as happened 20 years ago. But we'll have to move it.