Think More Thunk

A Wireless Joke

After digging to a depth of 100 meters last year, Japanese scientists found traces of copper wire dating back 1000 years and came to the conclusion that their ancestors already had a telephone network one thousand years ago.

Not to be outdone in the weeks that followed, Chinese scientists dug 200 meters and headlines in the Chinese papers read: "Chinese scientists have found traces of 2000 year old optical fibers and have concluded that their ancestors already had advanced high-tech digital telephone 1000 years earlier than the Japanese."

One week later, the Greek newspapers reported thefollowing: "After digging as deep as 800 meters, Greek scientists have found absolutely nothing."

They have concluded that 3000 years ago, their ancestors were already using wireless technology.

Certified Wireless USB (UWB) Forecast Divergence

There was a recent article in the EETimes about Artimi and their WiMedia-compliant dual-mode wireless USB / Bluetooth chip (the A-150). According to the article, Artimi will enter production at the end of 2007 with an initial unit price of $15. By 2009 the price may drop to $5, making it really competitive by the standards of today's marketplace. Artimi is an interesting company with a unique market focus and great potential.

However, the article offers an unattributed market forecast for Certified Wireless USB chips in 2009 of 200 million units. It is unclear if the forecast is an annual shipment forecast or if it is a cumulative one; either way the number is more than an order of magnitude above the current WTRS forecast for these chipsets.

There are companies, such as Wisair and Icron, which do have customers that are shipping product into the marketplace, however in the case of Wisair it seems that the chips are not yet Certified Wireless USB - compliant components. Wisair is in the business of developing rf component IP, however Icron is today selling its cable-free USB products which perhaps will be someday be included as part of the wireless USB specification. This is reasonable given the inevitable pace of a standards group as it lays down foundational rules, defines applications, and creates testing and interoperability compliance measures. With the exception of some software applications there are no products or markets that achieved 200 million units in its second, or even third, year of growth.

Put in perspective, 200 million units shipped means one device for every man and woman in the United States between the ages of 15 and 64. That is an impossible adoption to gain in just a couple years. There may be 2 billion USB peripherals in the world, but that level of adoption took more than 12 years, and really did not become a large market until the computer manufacturers were shipping USB in every PC and Laptop.

It is valuable to consider how market forecasts fit into the broader environment before blindly reusing them. In this way one avoids the inflated expectations that come with hyping a market before its time, and also the challenges faced when product forecasts do not materialize in two years. It is much easier to think about ones forecast first, rather than updating that resume two years from now.

Comparison of forecasts recently released:

• Spring 2007 - EETimes/Artimi - "Certified Wireless USB chips to reach 200 million units in 2009"
• Winter 2006 - iSuppli/Focus Enhancements - "160 million units in 2009"
• Winter 2006 - WTRS - "18 million Certified Wireless USB chipsets to ship in 2009"
• Summer 2006 - ABI - "UWB shipments to reach 300 million in 2011"
• Spring 2006 - Instat/Harris Wiltshire & Grannis LLP - "13 million UWB products in 2006, 190 million UWB products in 2009"

Is the IEEE a critical factor in developing industry standards?

One of the largely unforeseen consequences of developing UWB transceiver products outside the IEEE specification process is that many of the performance measurements for future RF chips are not discussed in a neutral environment; this includes not only the companies who will manufacture and sell the chipsets, but also universities and government research labs who all discuss the pitfalls and issues that can arise. Once UWB development left the IEEE and moved into a purely corporate arena, pitfalls and weaknesses were not openly addressed and thus are unlikely to be given the opportunity to be fixed. It appears that the development of MBOFDM UWB at this stage is largely political, and not an open, technical endeavor.

Therein lies the difference between the WiMedia Alliance and the Certified Wireless USB I/F, and the RF components based on the IEEE 802.15.4 specification which have been adopted by the ZigBee Alliance. The 'ZigBee' component-level products on the market today are differentiated, even though they all adhere to both the IEEE 802.15.4 specification and the ZigBee standard. Depending on a given application, one particular vendor's chips are better suited than another's, due to a tuning of various performance metrics to match that application's requirements. The components that are beginning to emerge from the UWB-focused standards groups are all focused on the same applications and the various performance characteristics are largely unavailable to the public. This means that the UWB components are not yet being sold except to very early customers, but it also means that their performance is probably wanting.

Thus, not only is UWB still a pre-emergent wireless protocol but once it does enter into the marketplace in reality, it is very likely that other technologies such as WLAN and Bluetooth will have taken its place in the applications currently being targeted by many of the UWB component companies. At that point it will be simply a question of whether the force and impetus represented by Intel can carry UWB into the marketplace. For those companies who placed all their bets on MB-OFDM taking over the market, it may be time to look to new wireless horizons.