Wi-Fi Alliance introduces low power, long range Wi-Fi HaLow | Hacker News Reader

Wi-Fi Alliance introduces low power, long range Wi-Fi HaLow

wi-fi.org

155 pointsdaegloe79 comments

There used to be a good wikipedia article about LoraWan, which is better than this (for IoT) in just about every way, but it has been replaced with this: https://en.wikipedia.org/wiki/LoRaWAN

LoraWan consumes power at a similar rate that BLE does, except it has a 20km range. You're talking about months (or even years, depending on your interval) of usefulness out of a CR2032 coincell battery.

Sidenote: the removal of the wikipedia page on this irks me more than it probably should. They're claiming a copyright problem because some of the text of the page was taken from the Lora Alliance's official website.

https://www.lora-alliance.org/What-Is-LoRa/Technology

> They're claiming a copyright problem because some of the text of the page was taken from the Lora Alliance's official website.

Stolen content (i.e. content not explicitly licensed as GFDL/CC by the author) is stolen content.

>Stolen content (i.e. content not explicitly licensed as GFDL/CC by the author) is stolen content.

Do you mean by wikipedia's definition of stolen? Because some things, like for instance: the description of Lora taken from the website that exists to provide descriptions of Lora to people, has an implicit usability.

Like I said, this irked me more than it probably should. I totally get sometimes stuff will get caught in the margins.

Wikipedia's mission is NOT about usability, but to provide an encyclopedia that can be used, edited, forked, reproduced EVERYWHERE without anyone having to fear lawyer attacks.

It's not a "free as in free beer" encyclopedia, it's a "free as in truly free" encyclopedia.

(Yes I know that this is only valid for the text content, images (e.g. logos) and screenshots are not free but used under different legal regimes in different countries)

> Wikipedia's mission is NOT about usability, but to provide an encyclopedia that can be used, edited, forked, reproduced EVERYWHERE without anyone having to fear lawyer attacks.

To that, I say [citation needed]. Where in their mission statement do they state this particular stance on lawyer attacks?

With all the fundraising activity, it should be about usability and put some aside for those instances.

> the description of Lora taken from the website that exists to provide descriptions of Lora to people, has an implicit usability

that's not how copyright law works. you may argue that it should work that way, but it does not work that way unless you petition Congress to modify it

Yeah, but it's a relatively safe bet that an organization would rather be represented by their own carefully-reviewed, professionally-crafted verbiage than one hastily thrown together by a volunteer editor, and thus would not pursue action against the publisher.

I'm not saying that Wikipedia should use content without the appropriate licensure, but in practice, copyright is violated as a matter of routine to provide some fairly basic services. Copyright law should be amended to prevent liability for conventional uses.

Multiply that "relatively safe bet" times the millions of times this would happen if Wikipedia relaxed their stance.

Like I said, I don't think it'd be a good idea for Wikipedia to start copying and pasting from random websites without the appropriate clearances. I do think it'd be a good idea to revise the law so that normal uses are not illegal anymore.

Still wouldn't change much because, thanks god, the US law is not global ^^

The U.S. has trade negotiations which tend to include mutual or compatible copyright law with most of the countries in the world.

Yes, they are among the biggest copyright aggressors of the world. Just like L Lessig, I care about common sense copyright policy, which forces me to care about reforming the U.S. campaign finance system so that congress can represent the people again.

Then the organization should have provided the text under an appropriate free license if they wanted it to be broadly reused. (Which is actually a good idea and more organizations should probably do for boilerplate they'd like others to cut and paste.)

Organizations would put the boiler plate under a "You can copy and paste, but no modifications, please." licence. Wikipedia would need special support (machine enforced or not) for blocks of text that can only be used or removed in their entirety, but not changed.

Come to think of it, that's not too different from how quotations work. So they might be able to deal with them.

I wonder how much text from a public (albeit copyrighted) source can be quoted under fair use, with attribution and all.

Here's what an unofficial essay on Wikipedia says and I think it sums it up sensibly:

> The copied material should not comprise a substantial portion of the work being quoted, and a longer quotation should not be used where a shorter quotation would express the same information. What constitutes a substantial portion depends on many factors, such as the length of the original work and how central the quoted text is to that work.

https://en.wikipedia.org/wiki/Wikipedia:Quotations#Copyright...

Wow, it claims 2-5km urban and up to 45km rural!

Important question though, is there an affordable module available (ideally also easy to use)? Sniffing around, all I can find is modules costing $45 or more, which kind of kills my enthusiasm for using it in personal projects.

Have a look for RFM98W (434Mhz) or RFM95W (868MHz) on aliexpress/ebay, they're around $14 there.

I know in the UK high altitude balloon community they've used them in flights - http://www.daveakerman.com/?p=1850

Although FSK works just as well in the case of balloons.

Ebay has them for less then $10, + $2 shipping (RFM98W (434Mhz) or RFM95W (868MHz))

LoraWan is super interesting for IOT but as far as I understand it acts more like a traditional radio transmitter? Great for collecting data from lots of discrete sensors but less for bi-directional data transfer

It's bidirectional, that's the whole point. If it was unidirectional it would be no better than standard radio.

I think there's also a bi-directional mode for it.

Isn't one for LoRaWAN specifically, but there are other options like Weightless-W (which doesn't have any hardware released for it yet).

The FAQ says bi-directional:

http://www.semtech.com/wireless-rf/lora/LoRa-FAQs.pdf

LoRa has long range but throughputs is extremely low. It is useful for some applications but not for all.

For most sensors throughput is not the problem -- it's distance, and power consumption.

Except that I can't buy it, it doesn't seem to have a stack, and doesn't have any open implementations to examine.

But, other than that, LoraWan is better.

Uh, yeah. Come back when LoraWan is shipping something other than vapor.

>Except that I can't buy it

Microchip sells radios that do LoRa: http://www.microchip.com/pagehandler/en-us/technology/person...

They're available on mouser for ~$17: http://www.mouser.com/ProductDetail/Microchip/RN2483-I-RM095...

>it doesn't seem to have a stack, and doesn't have any open implementations to examine.

Here's the source code for a development board with LoRa: http://ww1.microchip.com/downloads/en/DeviceDoc/LoRaMotePack...

So Sematech and Tata's LoraWan plan in India is vaporware? They claimed successful tests, and said their network is open to testing end-to-end by IoT developers.

This is a serious question, I just moved most of my stock portfolio into Sematech after that news. If Sematech's LoraWan is real, this is the best pureplay futurist stock I've seen since Tobii.

UWB was a wonderful technology and claimed successful tests and was open to developers. So was WiMAX (which actually did work). Where are they now?

Almost all the TI low-power RF stuff is real. Anything LoraWAN claims could have been done with these kinds of TI chips: http://www.ti.com/product/cc1125

So, why hasn't it? Mostly because these frequencies need big external antennas. TI has a very nice whitepaper about trying to create integrated PCB antennas for these lower frequencies and how much power you lose (answer--a whole lot). I don't think people are going to be very happy strapping a 6 inch whip antenna onto their tiny LED light.

And, these TI chips actually can run off a coin cell. Unlike the loraWAN stuff which flags Tx power consumption as up to 100mA. Coin cells are still ruled out.

That's not an improvement over, well, anything.

And, I note that Microchip is claiming to be shipping dev kits sometime this quarter. This is Microchip. I won't hold my breath, and I still mark this as vapor.

>I note that Microchip is claiming to be shipping dev kits sometime this quarter. I won't hold my breath, and I still mark this as vapor.

As of a few minutes ago seven in stock at mouser: http://www.mouser.com/ProductDetail/Microchip-Technology/DM1...

It looks neat, but it's hard not to be cynical about yet another competing standard for low-power "smart home" / IoT-type devices, just to muddy the ZigBee vs. Z-Wave vs. BT LE Longrange vs. conventional WiFi waters a bit more.

Not saying there's anything wrong with the protocol per se; it's probably perfectly fine. But the lack of widespread standardization in the 900MHz ISM low power space is really crippling what could otherwise be a robust market for smart devices. That's the key problem, rather than anything technical, and adding yet another option for hardware designers to choose from seems unlikely to improve things in the near future.

Zigbee is fairly awful. Most of that space wasn't designed for consumer products or multi vendor interoperability.

Being able to use the standard WiFi software & firmware stack but trading throughput for lower power and greater range has the potential to be a huge win. There are already many players who have WiFi silicon that it should be easy to adapt to this new protocol. Hopefully that will mean wide availability of solutions at competitive prices.

Relevant xkcd:

https://xkcd.com/927/

Are we really downvoting xkcd? I don't see anything wrong with a little bit of humour on HN.

That comic is a trope at this point. Any discussion of standards includes it.

The comment didn't add anything. If it was a more substantive comment and mentioned the comic you'd have a point.

But the comment might as well be a copy and paste job. It's not really adding to the discussion.

I mean just from the discussion and the letters XKCD I knew which comic it was.

https://xkcd.com/1053/

Fair enough, I'd have to agree.

Didn't vote, but if I could vote for "stay at 1 point" I'd do.

I mean, that comment isn't too bad but sometimes I (we?) get a sneaking suspicion that some of these (I didn't check the commenters comment history so I cannot say in this case) are attempts at "karma farming", which is annoying.

I agree, the wink hub already ships with something like 5 different radios. Also this doesn't look like it supports mesh networking. What advantages does it offer over z-wave or zigbee? And will it be better than Insteon's or Lurton's protocols?

It supports IP, which is a nice benefit for interoperability and standalone usage (it doesn't require a "smart" hub that understands all of the protocols).

But it does require a bunch of smart apps that can speak the language for each device you toss on the network.

LoRa mesh networking, not too many details but it appears to be possible: https://www.youtube.com/watch?v=EdejQkTdC6o

https://en.wikipedia.org/wiki/33-centimeter_band has more info on the 900MHz band they're using, which is an odd choice because it isn't recognized outside Region 2 (Americas).

900MHz seems a curious choice to me as an Australian, because I know my mobile phone uses 850 and 900MHz to talk to the mobile towers. Attempting to use 900MHz for wifi wouldn't get you very far due to all the interference.

Not to mention all the regulatory hurdles.

http://www.afar.net/tutorials/900-mhz-versus-2.4-ghz/

900Mhz does not buy much for distance, however it works better under difficult line-of-sight situations comparing to 2.4Ghz. For real long-distance 4xx/3xxMhz might be better, or even the frequency we used for FM/AM radios.

TI etc has those decent 9xx/4xx/3xxMhz chips for long time, it's just that they were not wifi-certified.

This is about minimizing power-per-range, not maximizing range itself. IoT devices have very local utility, one assumes.

I do like the idea of regular laptop-and-tablet Wifi having better range. But in dense environments, that might be an anti-feature: more interference from more (distant) neighbors. Suburban/rural might be the best use case.

I more would like the idea of laptops and smartphones being able to talk directly with low cost smart home stuff like your thermostat without having to have bridges or pay someone for access through some 'your data is now our data not yours' cloud services

Would this interfere with phones that operate on the 800MHz frequency?

Wi-Fi HaLow has a chance of success if it can be integrated into regular Wi-Fi access points.

LTE category-0 introduced with specification release 12 (and to be expanded in rel-13) is another competitor (other than LoraWan).

Here's more information: http://blog.3g4g.co.uk/2015/03/lte-category-0-low-power-m2m-...

I'd be happy if they just spec'd a 2mW WiFi, thanks. The fact that I can see my neighbor's WiFi is actually a hindrance, not a benefit, nowadays.

Output power is a setting in your wireless driver. Just set it to whatever is appropriate. (Getting your neighbor onboard is probably the bigger challenge.) Getting the power just right is a big part of planning large scale wifi installations. You could also put dampening (basically a resistor-like thing) in line with your antenna.

There's been several proposals how to do this automatically, which would do a lot for the congestion problem you mention, for the IEEE standardization process. But I have no idea if they made it into the newer standards. Unless you actively cooperate with your neighbors, it turns into a tragedy of the commons. The one with the power regulating equipment will have worse performance to start with, but if everyone has it then everyone benefits.

I'm somewhat concerned about LoRa and things like it for that reason. Having a transmitter emitting energy in your band is bad. But having a bunch of transmitters emitting coherent energy you're receiver can detect it worse. The former just cuts down your signal to noise. The latter directly limits your available bandwidth. Which is to say, if your radio detects the beginning of a weak packet from far away it's locked down for the duration. Can't transmit, can't receive another stronger packet that arrives slightly later.

LoraWan is inherently monolithic. There is only supposed to be one LoraWan provider, who acts as the router, and passes traffic to receiver according to their address -- just like a cell tower. You don't have a hundred cell towers, you have 3.

Not so. Sigfox has a single operator model: they're the only operator for their proprietary technology. But LoRa is multi-operator: anyone can buy LoRa base station and deploy their network (it's in unlicensed spectrum). And it's already happening this way. So several operators and even technologies will share the same limited unlicensed spectrum. Should be interesting when it scales...

It was designed to be "single operator", but that's not how it is shaking out. I think a trainwreck is coming, because if you have multiple LoRaWAN networks operating in the same area, all the packets are colliding at layer 2. They have to be thrown away by the server. And since the Semtech SX1301 baseband chip only has 8 demodulators, capacity quickly approaches zero, if there are multiple uncoordinated networks.

LoRaWAN is an okay solution for European Carrier networks where duty cycle limitations come into pay, but for the 915MHz band, it is not a good design.

Increasing WiFi range doesn't mean much for most consumers, but it's hugely important for deploying WiFi to large public spaces (city streets, offices, parks, etc.)

(On an unrelated note, the fact that you can see your neighbor's WiFi seems like one of the smallest imaginable hindrances and a terrible reason to reconsider WiFi innovation goals.)

How long is "long range" going to be ?

I hope we see a 8011.ah version of the ESP8266 soon. A solar powered GK-RadMon (https://sites.google.com/site/diygeigercounter/gk-radmon) would be the perfect device for this new standard.

Some very cool features included, I especially like using 900 MHz. I hope the WiFi Alliance doesn't pull the licensing crap that Bluetooth SIG has.

I am pretty sure there are many, if not most developed countries which has their GSM Network working on 900Mhz.

How is that going to work out?

So, can my Macbook connect to a HaLow router, or is this a new protocol that will require all new hardware across the board?

It looks like this runs in the 900MHz ISM (US) band, so we'll all need new hardware. The longer range is nice. Interference from other 900MHz transmitters (cordless house phones, baby monitors, etc.) may not be.

It seems everything old[1] is new again.

1. https://en.m.wikipedia.org/wiki/WaveLAN

I wonder if this will get any traction outside of the US - there are lots of places where 433MHz is the ISM band.

I'd hope no one's phone is running in 900MHz - what does DECT use in North America?

1.9 GHz

So signal drops every time you use the microwave... No thanks.

I thought microwaves (oven) mostly interfere with the 2.4 GHz band. That's why games of Starcraft always lag out when one of the players starts cooking a hot pocket. That's why it's preferable to use 5 GHz 802.11n or 802.11ac, unless you have walls that block the higher frequencies.

https://en.wikipedia.org/wiki/Microwave#Heating_and_power_ap...

Duh, you'd be correct. I had a brain-fart

This standard is targeted to low power consuming devices. So the question is more like "Will the new HaLow devices be able to connect to my home router". And looks like the answer is they won't.

I think after a while they will. Wifi chips supporting both current frequency bands cost pennies at the moment, if this standard gets some foothold it won't be too long before it's added in to generic chips.

Given that it runs at 900 MHz, it will require new hardware.

At a minimum, it will require an antenna tuned for the new band, in addition to whatever changes to the WiFi baseband processor are needed to go along with it.

Not exclusively, by the sound of it - the press release notes "Many devices that support Wi-Fi HaLow are expected to operate in 2.4 and 5 GHz as well as 900 MHz".

I interpreted the press release to indicate that WiFi HaLow will be 900 MHz only, but can coexist with traditional WiFi implementations on 2.4 GHz and 5 GHz. (The Wikipedia page seems to confirm this: https://en.wikipedia.org/wiki/IEEE_802.11ah) Just like when 5 GHz was added, this means getting a new WiFi card -- but that card could support all three bands at the same time.

So this would mean while you're near an AP you'll be on "classic" 2.4/5 GHz, but when you roam further away it would switch to 900 MHz "HaLow" for longer range.

Indeed, it does sound like that's the case. Thanks for the clarification.

Reply on news.ycombinator.com