802.11ah - HaLow

wired.com/2016/01/wifi-halow-internet-of-things/

A strange time to announce anything, early January.

Hugo, does Electric Imp have any reaction to this? I know it’s early days, but the prospects of high range, low speed, low power Wifi seems like a great fit for the Imp.

I note the comment about waiting until 2018 for certification. But, since 900MHz is an open band, individual vendors could still adopt it in an unofficial capacity until such time that they can put a genuine “HaLow” sticker on it. Right?

FWIW, “900MHz is an open band” depends entirely on what region you’re in. IIRC, in the UK (Europe as a whole?) 900MHz is licensed spectrum (current used for 2G)

802.11ah is not new - we’ve been talking with chip vendors about it for years now. I believe the announcement coincides with a formal spec ratification. None of the chip vendors we’ve spoken to have got products in the pipe for it yet, which doesn’t bode well for it being generally available in the next couple of years.

However, we’d absolutely love to support it, so when those chips do appear…

(plus, obviously, you need an AP which supports it too…)

The 802.11ah announcement uses the description “900MHz” in a pretty loose way (one significant figure, not even that in China). The actual frequency to be used in the UK is 863-868MHz.

Peter

It not really a large market for going only slightly further than wifi.

The need for longer range IoT is for the likes of parking meters in a city, beehives on rooves, cows in paddocks and trucks on the roads. 5G will likely include a channel for low power devices. The carrier prices for 4G IoT plans are already getting quite cheap, it’s the complex silicon and power draw at the device that’s stopping current mobile technology from being truly IoT capable, newer standards will address that.

I also doubt that the lower power will really be that beneficial. At the end of the day, in order to get to 1 year on 2xAA batteries running at VBAT (3.0V-2.0V) you have to get your average current draw down to 2700mAHr * 8760Hr/year = 0.3mA. Given that you still need a decent MCU to do the TCP/IP and Crypto that’s a pretty limited budget.

Making changes to the existing 2.4Ghz bands could likely achieve substantial reductions in radio power if APs had virtual private networks with no broadcast traffic and higher levels of DTIM. I suspect that is much harder to market though…