Over the years I have often had questions on network reliability of customer networks running the standard ZigBee software stack. I typically have answered this with anecdotal evidence and experience but not necessarily hard data. As ZigBee networks expand and are used for more applications, it becomes more important to document some typical behavior so engineers understand what to expect for performance. As new engineers start design cycles using ZigBee they question what level of reliability or what level of network operation can be expected.

ZigBee and WiFi Update

I had previously posted an entry on ZigBee and WiFi with some data on channel usage for WiFi from some of my travels. Having recently completed a trip to Asia I wanted to fill in some scans from this region since it was not previously represented.

I have now completed 10 larger scans in cities. This includes 3 in the US, 2 in Europe, 4 in Asia and 1 in Australia. The data is surprisingly consistent showing WiFi usage heavily targeted to channels 1, 6 and 11. The usage on these 3 channels varies from a low of 72% of the access points to up to 88%.

I have heard continuing discussion about code size and its growth over time in the ZigBee stack. As a stack provider, Ember has tracked code size over time since our conversion from a proprietary network stack to the ZigBee stack. Our initial stack was ZigBee 2004, then ZigBee 2006 and then ZigBee PRO. I thought I would share some of this data from the past 3 years.