News

Radiant collapse: not a threat to Mesh concept, says LocustWorld

Administration usually means "keeping going" rather than "winding up" - broadband wireless pioneer Radiant Networks is in administration, and has been since December when Ross Connock and Cedric Clapp went in to manage the struggling company. But commentators may have misjudged the importance of this. MeshWorks isn't anything like the MeshBox.

Both are used for providing broadband to remote rural communities, but that's the only similarity. The one is suitable for amateurs - the other is virtually impossible to install without major capital investment.

Professional installers do use the LocustWorld MeshBox - it is a very cheap PC with wireless, which acts as an Internet access point, and automatically links itself with any other MeshBox devices within range. But it can also be a DIY setup for a community at minimal cost.

By contrast, MeshWorks was a big, expensive system using proprietary technology, tall masts and (mostly) licensed wireless spectrum - requiring professional installation and management. Its inventor, Radiant Networks, was running several pilot projects in conjunction with the Government in places like Pontypridd, Wales - and to date, proving the viability of these projects has been challenging.

There are two Meshworks products; one running at the licensed 26/28 GHz frequency band, and aimed at operators like BT and NTL, for covering areas they couldn't get to with fibre. The other Radiant offering, using unlicensed spectra - 5.725-5.850GHz in the US, and 5.725-5.875 in European areas - was a recent add-in to the portfolio.

There is quite possibly still a future for Radiant, because big telcos need an alternative way to reach remote areas without having to dig up the roadside and countryside, installing new cables. Radiant offers a system of long-hop links, driving local nodes. It's essentially a way of getting T1/E1 access to a community, but offering far more bandwidth; and a way of distributing that bandwidth locally, without new cabling.

Its most recent announced contract was a commercial installation in California, last summer.

However, the way it uses mesh technology is completely different from the way Locustworld does. It has expensive, complex antennae, which have to be professionally installed and it is a conservative technology in terms of bandwidth. Where WiFi gets 50 megabits per second over a 5 GHz link (and sometimes twice that) Radiant offers users a peak data rate of 8 megabits, and a sustainable data rate of half that.

In the higher frequency, 25GHz band, Radiant speeds are more impressive, at 25 megabits peak, 10 megabits sustained - but this is not do-it-yourself territory. A licensed operator has to set it up and run it, with higher ongoing running costs. Planning permission is usually needed for the erections of the antennae; they can be very expensive, too.

By contrast, Locustworld technology is essentially self-configuring. Many meshes are entirely amateur. Professional ISPs do employ the technology, however, as (for example) in providing broadband to Newmarket.

Once the meshboxes are installed in the right locations, professional intervention is rarely required.

Finally, the meshbox approach provides what is effectively a neighbourhood LAN over which members of a small community can provide whatever services they like to each other - for zero cost.

The contrast has been graphically highlighted by the investment in Radiant and LocustWorld. Several tranches of multi-million pound backing have been poured into Radiant; LocustWorld has been financed almost entirely by "sweat equity" - hard work - by its founders, with a few tens of thousands of outside backing.

Some observers have suggested that the Radiant collapse implies a blow to the credibility of the mesh concept. In fact, the mesh component of Radiant was almost irrelevant to its business model, and it's the business model which let it down. Where there are a dozen or so Radiant installations world wide, and several of those are "pilot evaluation" projects, a Locustworld mesh goes live, on average, every day, somewhere around the globe.