In August, (see URL below) I wrote about the new Mesh Radio technology which is a cost-effective way of providing broadband services to customers. The basic idea is that there are no base stations and new network nodes only have to be able to see another node. Once a node is added it routes traffic for others in its neighbourhood. The fibre backbone is connected at convenient points within the mesh. The radio operates in the 28 GHz band and the Radiocommunications Agency is on the point of issuing the second batch of licences.

My article mentioned Radiant Networks, a Cambridge company, and described their equipment. Now, another company, CALY Networks, has come along, aiming to compete in this area. They are based in Sunnyvale, California and offer a system similar in concept but somewhat different in implementation. Radiant uses ATM as its native protocol within the mesh while CALY uses IP. Each company claims that its protocol is better but it really doesn't matter what they use internally as long as they can convey the protocols that the users want with appropriate quality of service. There seems to be no significant difference between them on that score.

One external difference is in the design of the antennas. Radiant has chosen to use up to four small horn antennas which can each be rotated to point at the next node in the mesh. CALY uses an arrangement called based on a Luneberg lens, which is a sphere made of different types of plastic. It has the property that a small antenna attached to the lens creates a beam in the diametrically opposite direction. There are sixteen feed points which means that the antenna can point in sixteen different directions in a 120 degree sector.

Again, there is little to choose between the two approaches; the CALY system has no moving parts, but the Radiant one has 360 degree coverage. Both can change beam direction very quickly.

Both companies are running trials with customers although Radiant is a little ahead at the moment.

It looks as though either system will work well. Unlike conventional radio networks, the mesh capacity and resilience increase with the number of customers. Mesh radio is one of those technologies that is so obviously a excellent idea that the market should be huge. The Strategis Group is forecasting a global market of $10bn by 2003 so there ought to be plenty of room for several players.

Mesh radio can provide data rates to the user of up to about 25 Mbit/s at costs that are comparable with xDSL. This should give new operators a chance to compete with the incumbents and avoid the hassle and questionable economics of unbundling of the local loop. The presence of competition in the supply of equipment will also be good for everyone. This is all great news for corporate and, later, residential customers. A single node on the roof of a shared building may be able to provide access for all the companies within; this could result in even better deals for the customer.

Keep a watch on the Radiocommunications Agency website and see who wins the licences for wireless access in your area - you may find a very good deal just around the corner.