Wireless will get tougher -but more interesting - Embedded.com

Wireless will get tougher –but more interesting

Unfortunately, while it may seem that going wireless should be getting easier, it's actually going to get tougher going forward for the non-RF expert as the FCC opens the door to new innovations.

Now more than ever, embedded systems designers are recognizing the value of wireless. From home light switches and entertainment systems to industrial controls to remote monitoring and communications, wireless is permeating an increasing variety of applications that might once have been thought of as either standalone or otherwise happily tethered to a wire of some sort.

While this trend should come as no surprise, given the ïexibility the wireless interface brings, there are many reasons it's taken so long for the trend to become the stampede it is today. Not least of these issues is the air interface's inherently vicarious link quality, added cost, and complexity at every stage of development, not to mention added FCC scrutinization and shielding issues. In addition, while choosing the correct wireless interface is difïcult enough, programming it to work to order has stalled many a project.

However, over the last 10 years, the development of the Internet has made ubiquitous communication almost mandatory in most of the world, while cellphone- and wireless LAN-type technologies have opened the floodgates to RF design expertise development and proliferation, thereby driving down RF front-end design costs, while at the same time making wired connections increasingly unattractive.

This combination of a need for ubiquitous connectivity (a need that's only going to increase as an extrapolation of Metcalfe's Law) and the falling cost of RF enablement has put embedded systems designers in a somewhat difïcult situation: They're being pressured to go wireless despite the fact that they're fully aware that though RF technologies may be proliferating, that doesn't make them easy to embed. From wide-area connections based on cellular technologies, to shorter-range Wi-Fi, Bluetooth, ZigBee, and whatever form ultrawideband signaling takes, wireless remains a mysterious and unforgiving force to be reckoned with for all but the experts. While the basic laws of wireless remain the same with respect to propagation, free-air power loss, absorption, reïection, fading, and so on—it's the basic laws' interactions with more nuanced effects such as intermodulation distortion and digital/RF coupling, as well as the varying behavior of components at different frequencies, that can tie up a design for weeks.

Unfortunately, while it may seem that going wireless should be getting easier, it's actually going to get tougher for the non-RF expert as the FCC opens up new bands and opens the door to industry innovation. One of the most exciting regulatory developments of late is the notice of proposed rulemaking (NPRM) on cognitive radio. This is a concept that will allow any radio to exist in any band so long as it has the ability to move rapidly if the licensed owner of that band comes online. It calls for new and unproven means of interference detection and avoidance, as well as elevated on-board intelligence, but if it comes to pass—as it's expected to—the technology will open up the airwaves to unheard-of levels of innovation.

In the meantime, what is a pressured embedded systems designer to do? First off, recognize that your value-add is not the wireless interface. Then, clearly define your connectivity requirements and bone up on the basic principles of wireless communications. Next, find a good consultant and/or silicon or ODM partner that can take you as painlessly as possible along the disaster-prone path to wireless enablement.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.