Open a laptop anywhere in central (even outer) London and the chances are that you will find four or five Wifi networks within range. There are only three non-overlapping sets of frequencies in the most-used 2.4GHz bands, which means that these networks are certain to be contending with each other - and this using spectrum already crowded with other applications, such as microwave ovens and Bluetooth.

Next-generation 11n Wifi technology could make the situation worse by extending the range and hogging bandwidth in the process known as channel bonding - equivalent to using two telephone lines instead of one to double throughput.

At the same time people will be encouraged to start using Wifi to stream high-definition video, meaning channels will be more intensively used and contention will get worse.  Wifi is not going to scale up well.

News of EC approval of ultrawideband (UWB) technology got submerged in Christmas seasonal spirit, but it could prove a good alternative (or complement) to Wifi. Of course it may turn out to have snags and it may, like Wifi, be hyped into uses for which it is not suitable.

But it looks promising. Firstly, it is fast: Wireless USB, which replaces the usual physical USB cable with a UWB link, is rated at 480Mbit/sec. Secondly, it is short range, designed for communication within rooms, which maximises the number of possible users in a given area (take note those Wifi vendors who boast of the unneighbourly reach of their products).

The short range is a problem, too, of course because UWB cannot by itself be used to distribute data around a home or office building. But Wifi is overused for this purpose, with the encouragement of companies that ought to know better. It should, where possible, be used to complement a physical network not replace it.

We will have to become ever more subtle in our use of spectrum as local wireless network links become even more widespread. As a general principle they should be of the shortest range and at lowest practical power; where a longer range link is necessary, a directional beam should be used to minimise interference with other networks.

UWB can be used for simple peer-to-peer applications such as dumping pictures from a camera to a PC or set-top-box. But it can be linked into a physical home network, and if it encourages more people to set one up, then so much the better.

It might also get companies thinking of ways to facilitate home use of Cat5 cable or fibre (see, for instance, the latest idea from Tenvera) instead of trying to kid people that Wifi is satisfactory alternative.

Ofcom does not seem to be falling over itself to get legislation framed for UWB, even though it has approved the technology in principle.  But while it is talking to parliamentarians, it would do well to discuss the establishment of some sort of best practice for networking homes – one that does not involve the promiscuous and anti-social use of scarce spectrum.