Broadening Roles for Wi-Fi

Some time ago I wrote a piece (although I can't remember where now) on the increasing number of radios (in terms of PHYs)  that might be put in handsets, and how we'd ultimately have to decide what subset of these would become the default in this application. Let's see, there's Wi-Fi, of course, in its various flavors and incarnations, Bluetooth (various radios and the protocol stack), the GSM/UMTS family, the CDMA family, Zigbee, WiBree, UWB, WiMAX, RFID, NFC, and a few more. Unless we're going to see a huge advance in the application of software defined radio (SDR) in handsets in the near term, which is highly unlikely, the requirement that this subset be defined is clear and, indeed, quite obvious. The chip guys are of course going to continue to push, and push hard, for their particular chosen technologies and implementations because they're obviously not going to make money conceding any point here. We saw this in spades with the Bluetooth guys, who succeeded in getting BT into essentially every handset even though it is used almost exclusively for headset applications.

Let's assume that a handset will have a wide-area radio of one form or another, and will also have Wi-Fi. The former is a really safe bet; the latter will see 50% penetration in handsets sold in 2012. Both of these radios serve difference applications, with some overlap, of course, increasing the popularity of mobile/mobile convergence, unless you believe that femtocells are all that's required, which is not going to be the case. But, assuming Wi-Fi in handsets prospers (and it will), what if we replace BT headsets with Wi-Fi headsets?

Even a few years ago, this would have been serious heresy. Wi-Fi was, BT proponents claimed, too big, too expensive, too power-hungry, and too slow (protocol-wise) for such an application. And yet I've seen several prototypes of Wi-Fi headsets over the years. Advances in Wi-Fi technology in fact now render the WLAN suitable not just for headsets, but for all WPAN applications. This is a broad set of possibilities, but think about all that USB or Firewire can do and you'll get the idea.

Case in point: I recently spoke with Ozmo Devices, a startup that just announced its presence and technology this morning. While they've yet to announce any products (which presumably will be through OEMs anyway), they did announce a low-power chip and associated drivers and protocols to make Wi-Fi PANs a reality. And I suppose the Bluetooth protocol stack could also run on such a configuration, so, while 3 Mbps (or less) Bluetooth is about as relevant today as 3 Mbps Ethernet, the broad range of Bluetooth application functionality should be available in Wi-Fi PAN applications as well.

It of course remains to be seen, but it may very well be that Wi-Fi PANs replace Bluetooth, UWB, and the assorted other contenders in the WPAN space today. Imagine using a Wi-Fi keyboard and mouse with your handset, and perhaps even using Wi-Fi to connect to a larger display as well. Do you still need to carry a notebook everywhere? Perhaps Wi-Fi will also play a role as the core technology in future active RFID implementations. I have always assumed that the economics and technologies of semiconductors would enable an ever-broadening range of possibilities for wireless of all forms. We have significant evidence of such being more true than ever in Ozmo Devices' announcement.