Three years ago, Bluetooth wireless connectivity was being hyped on the back of some frankly incredible market growth forecasts. Some predicted shipments of up to 500 million Bluetooth-enabled units in 2002.

The latest report from analyst company Frost & Sullivan estimates that only 9.23 million Bluetooth chipsets were sold in 2001. But it forecasts that 56 million will be sold this year, 155 million next year, and a billion in 2006.

These new forecasts look rather like the old ones, albeit with a shift of about 18 months to the right. But there are good reasons to believe that the new predictions are more accurate.

Bluetooth has been beset with all manner of difficulties, some real, some imaginary, in the past few years, most of which have now been resolved.

Bluetooth via 802.11b

First, there was the silly argument about the battle between Bluetooth and the Institute of Electrical & Electronics Engineers' 802.11b wireless local area networks, which has now thankfully gone away.

Then we had the worries about compatibility between different manufacturers' products. In fact, this one is still running but the Bluetooth Special Interest Group has done a lot to clear it up.

There will certainly be incompatibilities, but some will be intentional: do you really want your Bluetooth headset to communicate with your Bluetooth GPS receiver?

But, at last, most of what is now being written about Bluetooth is positive. Many people believe that the technology cannot succeed until Bluetooth chipsets cost less than $5. Frost & Sullivan reported that last year's average price was about $15.

But shipments are growing, causing prices to fall, so the cost might reach the $5 average soon.

If we are going to see 56 million units shipped this year we should, by now, be able to take a shrewd guess at which of the current crop of products will lead that growth.

The big sellers in recent years have been mobile phones and, even though sales are down, many of the new high-end handsets have Bluetooth capability.

The killer app

Bluetooth-enabled PC Cards for notebooks are still too expensive to ship in volume, so I think we have to look to mobile phones to lead the Bluetooth charge.

In fact, phones have a head start in terms of cost because the processors in phones can handle all the Bluetooth baseband stuff, which means that the only additional hardware they will need is a Bluetooth radio chip.

It looks as though all new smart phones will include Bluetooth. People who buy them will want to use Bluetooth with their notebook PCs and, once the mobile networks come up with sensible GPRS tariffs, there is likely to be healthy growth in the business market.

Consequently, it seems that the forecasts for Bluetooth growth are, at last, believable. It won't be long before Bluetooth will become so common that, like enhanced speech coding in GSM phones, nobody will even talk about it anymore.