WiFi on LoRaWAN bands (HaLow) offers good penetration and long range (802.11ah)

Рет қаралды 67,105

Ай бұрын

When I made my videos about LoRaWAN, some viewers asked if it would replace conventional Wi-Fi with a much bigger range. Unfortunately, this was not the case because LoRa has a huge range but a minimal throughput. It seems somebody listened to them and invented Wi-Fi on the same frequency band as LoRaWAN using Wi-Fi protocols. Cool. In this video, we will check it out to see if it can keep its promises, like going through walls and having a more than 1-kilometer range. We will compare it to standard Wi-Fi and learn that we must pay attention to staying legal.
My second channel: / hb9blawireless
Links:
HaLow extender WiFi: s.click.aliexpress.com/e/_DFt... or amzn.to/44T4rTu
HaLow extender RJ45:s.click.aliexpress.com/e/_DlY...
Mikrotik hap ac3: amzn.to/3yv6yRi
Mikrotik SXTsq 2.4GHz: amzn.to/3WPlIe9
Mikrotik SXTsq 5GHz: amzn.to/3WIZPx8
Mikrotik small 2.4GHz only router: amzn.to/44RKKeH
Patreon supporter companies:
www.welectron.com/
The links above are usually affiliate links that support the channel (at no additional cost to you).
Supporting Material and Blog Page: www.sensorsiot.org
GitHub: www.github.com/sensorsiot
My Patreon Page: / andreasspiess
Discord: / discord
If you want to support the channel, please use the links below to start your shopping. No additional charges for you, but I get a commission on your purchases to buy new stuff for the channel
My Amazon.com shop: www.amazon.com/shop/andreassp...
For Banggood bit.ly/2jAQEf4
For AliExpress: For AliExpress: bit.ly/3MtXUY8 (just go on from here to your product)
For Amazon US: www.amazon.com/shop/andreassp...
For Amazon.de: amzn.to/2r0ZCYI
For Amazon UK: amzn.to/2mxBaJf
For ebay.com: ebay.to/2DuYXBp
/ spiessa
Please do not try to email me. This communication channel is reserved for my primary job
As an Amazon Associate, I earn from qualifying purchases
#no#midroll#ads

Пікірлер: 364

@user-ov6gv3xe4z Ай бұрын

i want to dedicate my graduation to mr Andreas, thanks to him i built a big IoT project.

@Artichoked1 Ай бұрын

what did you build?

@user-ov6gv3xe4z Ай бұрын

@@Artichoked1 the graduation project consists of a set of modules that gets agriculture data and sends it via lora and controls irrigation as well, i've been also looking for a way to send/receive data without using wifi or any other transmission method that is not guarenteed to be available in certain places, the LoRaWAN did a great job and it served the command better than any other method and with this channel i learned everything.

@ladedk Ай бұрын

Cool, congrats! Andreas is a super skilled educator and communicator.

@ww07ff Ай бұрын

@@user-ov6gv3xe4z Congrats man!!!

@Artichoked1 Ай бұрын

@@user-ov6gv3xe4z Awesome! I'm thinking of doing something very similar for my year 12 project next year. I'm in an area where grape growing for winemaking is a huge thing so I'm designing lora devices that can monitor weather and soil data for viticulturists.

@jean-marclugrin1902 Ай бұрын

Thanks. The reinforcement steel in concrete is (normally) grounded, an additional reason why it is very difficult to go through it.

@AndreasSpiess Ай бұрын

Great point!

@digus Ай бұрын

I never considered this, that would explain a lot.

@chargehanger Ай бұрын

At 868 MHz, Grounding does not mean what you think it means.

@Ariccio123 Ай бұрын

Well, yes, the concrete is usually in the ground 😅 Watch some of mike Holt's videos!

@Uncle-Duncan-Shack Ай бұрын

Hi Andreas, The analogy that radio behaves more like light as the wavelength decreases is one very sensible explanation of what happens, anyone can understand that, I like! As always, I enjoyed the video. I enjoy the talks on new stuff and radio devices which is always educational to me as I have a lay understanding of rf. I need a basement now 🙂 Regards, Duncan

@SoundOfYourDestiny Ай бұрын

It's not quite complete, though. The full comparison is that higher-frequency light bends more as it transits different media, just as higher-frequency radio does. It explains why 5G cellular and 5 GHz Wi-Fi are often useless (as he points out). This depiction is correct; note that the bluer (higher-frequency) light bends more: upload.wikimedia.org/wikipedia/en/3/3b/Dark_Side_of_the_Moon.png

@AndreasSpiess Ай бұрын

Thank you! But: The roof is much more important than the basement if you want to play with RF ;-)

@Uncle-Duncan-Shack Ай бұрын

@@AndreasSpiess yes, you are correct

@johnwest7993 Ай бұрын

@@AndreasSpiess, I actually had a better signal to noise ratio for my 60 kHz WWVB receiver when I put the loop antenna in the basement of the old 3 story building my shop was in than when I tried it on the roof. There were AC power-lines running along the roof-line, but the building was so old that there was no AC power in the basement, so I had better reception down there.

@SoundOfYourDestiny Ай бұрын

@@Uncle-Duncan-Shack Odd, my comment seems to have disappeared.

@user-pk4hn1uz1k Ай бұрын

Your LoRa videos have been super helpful as primer, the GPS video was also very interesting

@AndreasSpiess Ай бұрын

Thank you!

@batterynerd8779 Ай бұрын

Hello, i am an ongoing electronics engineer. Right now i am on a project that uses LoRa as communication. Your videos on this topic have been a great inspiration! Grüessli usem Aargau :)

@AndreasSpiess Ай бұрын

Good decision to becom an EE !

@BenJefferyCanada Ай бұрын

Glad to see you covering HaLow! When I rediscovered 802.11ah last year and saw that you didn't have one covering the standard, it inspired me to make my video.

@AndreasSpiess Ай бұрын

And you video was pretty successful! As said: I think, HaLow is mor suited for your region.

@johnwuethrich4196 Ай бұрын

Thanks for your videos! Always awesome :)

@AndreasSpiess Ай бұрын

You are welcome!

@coppurt Ай бұрын

I didn’t even know we had our own men in black in Europe, this looks like a cool trick to summon them 🙏🏻

@AndreasSpiess Ай бұрын

With the help of Dall-e

@shamasis.bhattacharya Ай бұрын

Thanks for making the video on this topic.

@AndreasSpiess Ай бұрын

My pleasure!

@trentbrown3714 Ай бұрын

Andreas, thank you for the great HaLow overview, it is a very interesting technology, hope it gets more attention!

@AndreasSpiess Ай бұрын

You are welcome! I assume it will be more used in the US because there, its value is bigger. But I will keep it around for ad-hoc problems.

@MikeKranidis Ай бұрын

Super informative, thanks.

@AndreasSpiess Ай бұрын

My pleasure!

@mousbleu Ай бұрын

Thanks a lot, another super interesting video :)

@AndreasSpiess Ай бұрын

Glad you enjoyed it!

@plasmamac Ай бұрын

Danke Andreas. Cool wie immer 😀

@AndreasSpiess Ай бұрын

Danke!

@Chris-ut6eq Ай бұрын

Very nice info, quite useful!

@AndreasSpiess Ай бұрын

Glad you think so!

@Hex-Mas Ай бұрын

Well this channel has done it for me.

@AndreasSpiess Ай бұрын

Is this good or bad?

@Hex-Mas Ай бұрын

@@AndreasSpiess 👍

@ericblenner-hassett3945 Ай бұрын

It was interesting. The board reminded me of an idea I had when the ESP-32 was new of creating a wi-fi/Lora switch where the ESP32s would do a translation of wi-fi to Lora and connect devices to a " normal " router. With the newer modules, it would be interesting to see if WEP2 encription is now possible with open source coding.

@AndreasSpiess Ай бұрын

The "real" LoRa devices have a much lower speed. So they are not made for WiFi, I think.

@antibrevity Ай бұрын

Excellent. I would actually have a use case for this, so I plan to learn more.

@AndreasSpiess Ай бұрын

Cool!

@holgerschurig4430 Ай бұрын

Many years ago, when I joined a provider for electronic equipment for logistics, they still had sub-GHz radio equipment that was used for TCP/IP. Terminals mounted onto fork lifts were able to use it,. Speed was certainly below even 802.11b WIFI. But back then a lot of the logistics software used either Telnet or 5250 protocol. So quite a small amount of data was exchanged to update screens. I know that they had HP spectrum analyzer going to 1 GHz that they "abused" to check radio ranges in the warehouses. Personally, I never worked with this equipment. However... if you say that someone invented sub-GHz WIFI, then yes, this was certainly done. About 20 years ago! 🙂

@AndreasSpiess Ай бұрын

I did not know that they already used standard IP. Usually, these links used propretary protocols. So thanks for the hint.

@midnightrambler4468 Ай бұрын

great video, love it.

@AndreasSpiess Ай бұрын

Thank you!

@TymexComputing Ай бұрын

Thanks you Andreas Spiess for new video - will watch it in the evening! Andreas in his hat is both 4 non blondes and for blondes :)

@gustavrsh Ай бұрын

Great video! Suggestion: you can use iperf3 to test network speeds without being hindered by your ISP speeds :)

@AndreasSpiess Ай бұрын

I use it on Linux. For my Windows PC the ISP thingy was easier ;-)

@tomfahey2823 Ай бұрын

@@AndreasSpiess Windows Subsytem for Linux is a good option for running iPerf3 on Windows! (apprently the native Windows compiled binaries have limited support for Windows network API calls and suffer from a performance penalty, due to having to go through an emulation layer)

@avejst Ай бұрын

Another impressive test 🙂 Thanks for sharing your experiences with all of us 🙂

@AndreasSpiess Ай бұрын

My pleasure!

@jwshields Ай бұрын

Wonderful video, thank you! And while I know you're not a "Networking" channel, I do appreciate the more recent appearances of Mikrotik devices! On that note, have you ever looked at the 60GHz range of devices? They have some really nice PTP & PTMP transceivers. I own a set of Wireless Wire Cubes, and it forms a symmetrical 1G link across my property. Higher end 60Ghz devices can allow multi-KM links (provided there's line of sight) - Pretty neat stuff imo

@AndreasSpiess Ай бұрын

I even made a video about them 😉

@IndependentNewsMedia Ай бұрын

Nice overview video, God bless.

@AndreasSpiess Ай бұрын

Thank you!

@Lorem_the_Ipsum Ай бұрын

I use the same thing with another brand, i opened it up and removed the coin of metal that made up 90% of the weight. Probably lead. I then cut out some of the inside plasics, added a battery and a switch to make it more mobile. Highest range i got to work was 250m at around 0,5mbit speed. Works through 3 to 4 walls of reinforced concrete. I use this in industial buildings and construction sites to connect to some controllers while checking sensors and devices

@AndreasSpiess Ай бұрын

Thank you for sharing your experience. Here there was no lead inside... Good idea with the battieries (if they do not interfere too much with the antennas)

@Lorem_the_Ipsum Ай бұрын

@@AndreasSpiess I tested the range before and after I added the batterie, I didn't see any difference. Batterie runtime is ~60 hours with my 18,5wh cell.

@john-r-edge Ай бұрын

I would be cautious about use of point to point radio comms over water. In a basic setup you may encounter "multipath fading" when the remote antenna gets one signal directly, and one which has reflected off the water. If these arrive half a wavelength out of phase, the two signals wipe each other out. To understand this requires a lot of factors to consider, wavelength, height above water, air temperature which changes humidity and so the refractive index of air. A well engineered setup may have two independent links with "space and frequency diversity" where one link will be unaffected if the other fades - and a switchover device to select the best. I worked on telecomms for a company with platforms quite close to shore. Our consultant had worked all over the Pacific and had hprror stories about radio links installed which never worked or fauled frequently because of these phenomena. Hence we put a lot of work in ensuring such problems did not affect us.

@AndreasSpiess Ай бұрын

I agree. This is probably the reason why these guys were not successful with conventional Wi-Fi and changed to HaLow.

@krishnaprasad8978 Ай бұрын

Thank you Andreas for the exceptional clarity in explaining the topic you choose!. I am wondering if you could do an episode on Single Pair Ethernet - I am trying to the PSE prototype working for a regular use case of powering and measuring sensors and MCU's. I have tried both the Beagle Play with TI chip and the board from Analog Devices without much of a success. Maybe a Swiss intervention, could be the solution!

@AndreasSpiess Ай бұрын

So far, I did not see single pair Ethernet in the Maker community. So it is probably not a good topic for my channel. It is more used in industrial setups.

@richard_wenner Ай бұрын

Just knew you were going to mention password and 'Breaking the code' this week whoo ho ho!

@AndreasSpiess Ай бұрын

You should know that I am an optimist and therefore not very much into security. But this one was obvious ;-)

@ErikThiart Ай бұрын

It will be a game changer if MikroTik joins

@AndreasSpiess Ай бұрын

We will see...

@carlotube111 18 күн бұрын

Hi Andreas, always very interesting and clear videos! Would you have info on how to configure the frequency, bandwidth power with AT commands?

@AndreasSpiess 13 күн бұрын

I have some. But they did not work on my version

@carlotube111 12 күн бұрын

@@AndreasSpiess Pity!, I will keep trying to find out I will let you know. Thank you, Carlo

@peterblackmore7560 Ай бұрын

Very interesting!

@AndreasSpiess Ай бұрын

Glad you think so!

@yancymuu4977 Ай бұрын

I have been working on a LoRa long distance low data rate link to connect an RTK GPS base station to my computer. This is turning out to be a complicated project at least for me. This HaLow might be an easy solution. Thanks for great video.

@helmuthbecker7636 Ай бұрын

HaLow uses IP protocol, it turns everything much more easy in projects.

@AndreasSpiess Ай бұрын

For sure an alternative!

@antoineroquentin2297 Ай бұрын

Thank you for the in-depth analysis. Do the devices respect the CH/EU duty cycle limitation?

@AndreasSpiess Ай бұрын

I did not find any duty cicly limitations in the 802.11ah specs. But maybe I did not properly search

@scotty3114 Ай бұрын

Thanks!

@AndreasSpiess Ай бұрын

Thank you for your support!

@PhG1961 Ай бұрын

Again a very interesting topic, very well explained.

@AndreasSpiess Ай бұрын

Thank you! I hope you had a pleasant long weekend ;-)

@PhG1961 Ай бұрын

@@AndreasSpiess Excellent weekend... 4 days in my workshop!

@AndreasSpiess Ай бұрын

@@PhG1961 Cool!

@sharg0 Ай бұрын

Regarding CE marking and conformity, if I'm not mistaken you are technically the importer and thus the one responsible to take care of that (aka tons of paperwork)! I have no idea if any private person have been hold to this but it's a risk if it interferes and gets reported. My personal opinion is that the laws should be updated so the large mail-order companies (in particular) that actively markets products to Europe should be required to take this responsibility when dealing with individuals.

@AndreasSpiess Ай бұрын

I am no lawyer, but I think, the manufacturers can go for the CE label, too.

@sharg0 Ай бұрын

Can go for, yes but what I've been taught is that the one that is responsible is A: The manufacturer if the product is made within the CE-countries. B: The importer if it's manufactured elsewhere.

@eliluong Ай бұрын

excellent overview! Would the HaLow extender RJ45 work with VLANs?

@AndreasSpiess Ай бұрын

I did not test it (I have no such device).

@BR0KK85 Ай бұрын

I had the men in gray at home once beacuse i bought an US ordering system wich uses frequencies that are not allowed in germany. I was a bit "grün hinter den Ohren" and they found me. They gave me a stern talking to and explained how they "found" me. They monitor frequencies with ther unsuspecting vans with huge antennae on top ....

@thetinkerist Ай бұрын

Did you get fined?

@BR0KK85 Ай бұрын

@@thetinkerist no luckily not but I got a a really stern talking to by them. They were nice overall because 1. I was nice to them 2. I was clueless (it showed!) what I did wrong and I really wanted to know why and how.

@Richardincancale Ай бұрын

I’m guessing you were a bit ‘green behind the ears’? Same expression in English! :-)

@AndreasSpiess Ай бұрын

Thank you for sharing. That is why I warned in the video ;-)

@IceArdor Ай бұрын

The 2.4GHz and 5.8GHz network throughputs might have been low not because of a low MCS or poor SNR, but could have been due to channel congestion. 802.11 unlicensed transmissions require listen before talk, so your devices may have spent a lot of time waiting for the channel to be clear before they were able to transmit. Depending on the protocol used in the higher layers of the OSI stack, the max delay/timeout without receiving an acknowledge may require retransmission, further reducing your data throughput (depending on which throughput you’re measuring). Consider measuring at night time when there’s less wifi traffic to see if there’s a difference. If you have some low-level diagnostic data, that might give some insight into where the link is falling short.

@AndreasSpiess Ай бұрын

Well possible, because I have a lot of 2.4 GHz gear around my home. But this might be valid for other viewers of this channel, too ;-)

@klassichd10 Ай бұрын

Thank you for your effort and this very interesting insight into this technology. .

@AndreasSpiess Ай бұрын

You are welcome!

@steubens7 Ай бұрын

the 5.8ghz performance might be due to higher mcs / modulations in newer specifications, in 802.11ax those are also available to the 2.4ghz band. wavemon in linux can show you the mcs & link bitrate choices (there's some android apps too) apples to apples comparison can be done on most router firmwares by setting '802.11 only' station mode. i'm almost willing to bet that a newer router will do 2.4 a lot better and an older would do 5.8 worse

@steubens7 Ай бұрын

there's also a lot more spectrum space to fill with error correction in the wider channels on 5.8, most firmware will let you change those as well (eg. restrict 2.4 and 5.8 to 40 or 20mhz channels)

@AndreasSpiess Ай бұрын

You are right, I could have reduced the bandwidth of the 2.4MHz link (go back to the old protocols). However I wanted to keep the standard configuration.

@berlinberlin4246 Ай бұрын

@@steubens7 can you name the android apps please?

@mph8759 Ай бұрын

What is the wifi scanner program you use? Thank you for the informational video, with practical test and benchmark.

@mdevries8495 Ай бұрын

If you're running Linux you can just use the 'ip' commandline interface to scan for wifi signals and get information. On windows I used to use Inssider to get info, but I think they dropped their free tier.

@AndreasSpiess Ай бұрын

It is built into the Mikrotik RouterOS

@johnwest7993 Ай бұрын

Thanks for the analysis and comparisons. Last I checked I'm a USA Amateur Extra license holder, so running a few Watts (I think I have a 10 W power amp and some receive amps for that freq range.) Those and a couple of beams on the 900 MHz band sound like fun, but I think I'd have to do so without encryption. I'll need to reread the FCC regulations before I put these on the ham bands with power and directional antennas, but it sure would be fun to do some experiments from a ham friend's place at 9000 feet up the side of the mountain here in Colorado.

@AndreasSpiess Ай бұрын

Your mountains are also fun without Wi-Fi. I still remember my 4x4 trips to the Engineer Pass and some others around Silverton ;-)

@Mr.Leeroy Ай бұрын

WPA2 is not terribly secure by today's standards.

@StuffJason437 Ай бұрын

You'll be ok with long password.

@AndreasSpiess Ай бұрын

Thank you for the additional info!

@IceAce1 Ай бұрын

Notably. the recent CVE-2023-52424 does affect WPA3, but not WPA2.

@tarakivu8861 Ай бұрын

@@IceAce1 And to be fair, SSID confusion attack is not much worse than auto-connecting to insecure networks. If your program is confused by the SSID name and doesnt have proper Firewall rules in place against leaks and unwanted access.. you are not doing a good job in the first place.

@IceAce1 Ай бұрын

@@tarakivu8861 Well, it's a protocol attack enabling a rogue MITM access point, which can subsequently read transmitted traffic. A firewall on your end points cannot help against that really. By the way, WPA1 (which is used here) is also considered safe wrt to this CVE.

@eon6008 Ай бұрын

Hi, your video comes at just the right moment. For the last few weeks I've been looking for ways to establish a data connection in caves. As cave explorers, we always have the problem of being able to communicate over long distances underground. At the very least, a person has to be sent to communicate, which can take hours or even days. The idea is a kind of mesh system that works on battery power for at least a few days. It must be possible to transmit at least text messages and images. This may also take a moment. The question is on which system should this idea be based? Lorawan is not the right one as far as I know, otherwise I only know espNow. Do you have any ideas? Thanks for any feedback😁

@tarakivu8861 Ай бұрын

Why not TTGo Lora boards with relays along the path? Anything with higher frequency wont work well in confined spaces anyways. And even the lower ones wont penetrate walls in a cave very deep.

@moki5796 Ай бұрын

How far through solid rock would the signals have to travel? Do normal hand radios work down there? If they don't, neither will LoRa or 802.11ah.

@AndreasSpiess Ай бұрын

I would look at the Meshtastic project for such a use case.

@kworgz Ай бұрын

Meshtastic using lower powered nrf microcontrollers. This will solve your text messaging issue but will not solve sending images. Luckily, these devices are relatively cheap. If you use directional antennas and can place enough nodes at chokepoints, HaLow should work. It would just be exponentially more expensive.

@richardparsons6156 Ай бұрын

I question its usage of the 868Mhz band and airtime as per the legal requirements.

@AndreasSpiess Ай бұрын

This is what I thought, too. But 802.11ah seems to be legal. And I did not find any duty cycle restrictions.

@remotepeak Ай бұрын

I agree with you. Norm: EN 300 220 point 5.21.3.1. says about maximum allowed cumulative on time usage: 100s on each 1h period. Maybe there is something more in norms, but since I remember, this band wasn't allowed to continous use.

@SurfinScientist Ай бұрын

@@remotepeak The duty cycle limitation may depend on whether Carrier Sense is used or not. In Japan the duty cycle limit is 10% with Carrier Sense and 1% without Carrier Sense. It may also depend on whether frequency hopping is used. I don't know the regulations in Europe, but I suspect that there are similar restrictions.

@saxpert Ай бұрын

Duty cycle depends on the exact frequency. A constant tx is illegal anyway in Germany, 869.5 allows 10% and 500mW.

Ай бұрын

At least in Germany, there are no such duty cycle restrictions on HaLow. Edit: There needs to be some sort of AGC, however. Otherwise, the maximum duty cycle must be under 10%.

@logmeindog Ай бұрын

Lack of proper encryption means an automatic NO.

@AndreasSpiess Ай бұрын

That is why I mentioned it. So you can decide.

@roboman2444 Ай бұрын

99.9% of the time, my phone doesn't need more than a few kbit/s of bandwidth. Being able to use chat apps, get emails, or get traffic info in maps without having to turn on "data" would be fantastic.

@AndreasSpiess Ай бұрын

Indeed, 2Mbps is ok for a casual user...

@thetinkerist Ай бұрын

Thanks for all your work, can I kindly request a 2.4 Ghz test while it is raining? Water is supposed to dampen the signal a lot esp. for 2.4, but I've never conducted a test for this.

@AndreasSpiess Ай бұрын

It does not have a big influence. My 2.4GHz uplink reaches a geostationary satellite at 36000km distance also during rain.

@user-ng2ly6nf1c Ай бұрын

Thx for sharing that test!

@AndreasSpiess Ай бұрын

My pleasure!

@o0julek0o Ай бұрын

I was waiting for you to cover this.

@AndreasSpiess Ай бұрын

So you have it ;-)

@JamesHalfHorse Ай бұрын

900mhz wifi was used in the US mostly by wireless internet providers for point to point but mostly phased out for 2.4 and 5ghz (and other licensed bands) for more bandwidth. They stopped making the 900mhz and it's hard to find now because it's scooped up for low bandwidth long range bad LOS stuff cctv and the like. Hopefully this will be a cheap replacement for it. 2.4 and 5 Wifi around here is tricky due to being in the mountains.

@AndreasSpiess Ай бұрын

Thank you for the additional info. But I am not sure if it helps a lot with mountains. It still depends a lot on line of sight.

@JamesHalfHorse Ай бұрын

@@AndreasSpiess Eh... technically but a lot of the problems with the terrain was all the trees on tops of those hills and mountains and it did help with that. Any LOS calculation here add 60-80ft on your overall height for trees on top of whatever the site calculators give you and even that is not enough in some cases. I think back then we were using Breezecom 2.4 and 900mhz Motorola Canopy systems. 2.4 you practically had to be able to see the tower from where you were. The 900mhz was a lot more forgiving but due to bandwidth limitations it just kinda fell out of use and nobody makes them anymore. The ubiquiti 900mhz stuff was awesome and gets snatched up quick when it goes on ebay so I don't know why they don't make it anymore. Before we shut down we were using Navini which used some early beam forming to the CPEs to help with line of sight. I think Cisco bought them out for the tech.

@skrywenko6596 Ай бұрын

@@JamesHalfHorse yeah I maintained a community base ISP we had 900mz 2.4 ghz and 5ghz , the 900mhz stuff could easily punch through a mile of trees and if you where in lightly wooded area of the tower or if you had clear line of site for few miles you did not even need to put an antenna outside just the client radio and the indoor antenna ( like a router ) . . the 2.4 not only did you need line of site it had to be fairly high above the treeline so not to cut off the bottom the signal wave. 5 ghz stuff was even worse. as fog and cold air layering can be a huge problem

@JamesHalfHorse Ай бұрын

@@skrywenko6596 Yeah that's pretty much been my experience. I don't know why 900 wifi went away like it did as it's still so useful in terrain like ours. We still use licensed 900mhz analog links for the studio to transmitter links for our radio stations not yet on internet based encoders. Rock solid. The 2.4 and 5 stuff is tricky for a wisp. Mine had to put me from 60ft up to nearly 80ft total to get me off a 2.4 local tower and shoot me back to their main tower the fiber comes into instead because I work remotely sometimes for them. The calculator says it's clear at 60 but if you look at it there are outer lines of the signal that hit the trees that will wreck it. Also a lot of trouble back in the day when we first were doing 2.4 especially with pine trees until the company told us pine needles were just about right size to act like antennas and scatter the signal especially when wet. You would get good signal even mostly line of sight then when it rained it would go to hell. The 900mhz stuff just worked but as a wifi option its all but gone.

@jmr Ай бұрын

I've been waiting for HaLow for years. My hope is that it will eventually become part of regular WiFi routers and be practical for IOT devices. I had a WiFi like device that operates in the 900mhz range many years ago. It was a one to one device. One end connected to my dial up and the other my laptop. 😂

@tarakivu8861 Ай бұрын

I dont see the need to have yet another Wifi and IP-based system for many IOT devices that other protocols like Thread and LoraWan already solve. Many operate on an ad-hoc basis anyways, so they wake up, send data, or only wait for events to happen. Anything that needs a more stable connection probably also needs a higher bandwidth anyways.

@jmr Ай бұрын

@@tarakivu8861 I see what you are saying but this is a good system for devices like cameras that might be too far away for WiFi but need more bandwidth then Lora, Thread, or ZigBee. More bands are always nice too in crowded areas.

@AndreasSpiess Ай бұрын

I think that router manufacturers go more in the oposite direction: More speed and therefore higher frequencies. So this band will be a specialty also in hte future.

@mosi44 Ай бұрын

Speaking of regulation in Switzerland: In addition to the frequency band, there are also limits for the duty cycle. Are these requirements also met?

@AndreasSpiess Ай бұрын

I did not see any regulations in the 802.11ah concerning duty cycle. Probably it would not make sense…

@silverian Ай бұрын

Thank You for HaLow information!

@AndreasSpiess Ай бұрын

My pleasure!

@Sh4dowHunter42 Ай бұрын

Yes! YES! YEAH! WOOHOO!

@yt3dkraft Ай бұрын

I think it should be made clear that comparing an omnidirectional link to a link with directional antennas are two different applications, so it is a bit like comparing apples and oranges.

@AndreasSpiess Ай бұрын

I thouhgt that I mentioned this fact in the video.

@SightsToKeepInSight Ай бұрын

Very interesting topic. Thank you again

@AndreasSpiess Ай бұрын

You are welcome!

@tonyh6309 Ай бұрын

[EDIT: simplified] Probably a major reason the 5.8G link was so much better than the 2.4G link was because, as you noted, the former has an antenna with 6dB higher gain. The regulations limit the ERP so using a higher gain antenna means the transmit power has to reduced. The 2.4G link had 5dBm greater TX power (30dBm v 25dBm) but a 6dB lower gain (10dBi v 16dBi) antenna. The extra 5.8G's extra 1dBm is likely immaterial, so both links have similar TX power. However, the *BIG* difference comes from the fact that the receiver also benefits from the antenna gain so the 5.8G receiver could have been receiving 4x, (6dB) more signal power than the 2.4G link with potentially 4x the range. I say 'could' because antenna radiation patterns can be 'interesting' and require careful alignment at both ends for best results. Also radio propagation conditions due to the ground/buildings etc. reflections, water vapour and so on may have had a much higher impact. The ground reflection interfering with the direct path can have a big effect at very specific ranges and moving the antenna a meter or so closer or further away could have changed the results drastically. I believe this is particularly true when either, or both antenna are low down. Up to a point, when ERP (or EIRP) are limited, and you can use highly directional antennae, then you are better off using higher gain antennae with lower TX power. It also saves a lot of power consumption. Another benefit is that interference from sources that are not directly in line with two stations can be significantly reduced due to the lower gain of the antenna to off-axis signals. Of course, as always, there are many other factors and tradeoffs to consider.

@AndreasSpiess Ай бұрын

I thought about that. But the difference was too big for me. These few dBs are not enough explanation (remember, 5.8GHz delivered full speed, and the free room loss is higher on 5.8)

@Mystik314 Ай бұрын

As a network engineer, I just want to issue a small correction. You keep saying 5.8Ghz, which is an ISM band. WiFi (802.11) uses the UNII bands for 5Ghz, which extend all the way from 5.1 to 5.9 Ghz. so 5.8Ghz is included, but WiFi is not limited to only 5.8Ghz operation.

@AndreasSpiess Ай бұрын

You are right. I I would say 5G, people would mix it with another standard…

@chris-tal Ай бұрын

What about ETSI/CEPT regulations on duty cycle? Does the "Listen Before Talk" rule (like using CSMA/CA) and FHSS make this legal without using a low enough duty cycle or none at all?

@AndreasSpiess Ай бұрын

It seems there is a special regulation for 802.11ah. I did not find duty cycle limitations.

@PhilippBlum Ай бұрын

@@AndreasSpiess As far as I remember the standard is explicitly designed for 915 Mhz NA bands in mind. So, these regulations don't apply to Europe. Or did I miss something? Do you have some references? I bought some of those and did some research. I was pretty sure they were completely illegal due to the strict duty cycles in Europe.

@AndreasSpiess Ай бұрын

@@PhilippBlum I checked in Switzerland, and it is on the Wi-Fi standards list. Usually, we adopt EU regulations.

@PhilippBlum Ай бұрын

@@AndreasSpiess KZfaq may deleted my comment. I sent you a dm on twitter with a link :)

@axelhertwig8258 Ай бұрын

I also tried to find respective exceptions in latest ETSI specifications with no luck. Standard is 1% duty cycle. In some discussion papers I see 2.8% and 10% for 863-868MHz if polite access is used, but I could not find anything official. Yet 10% seems to be insufficient for many use cases.

@anon2030 Ай бұрын

It would be interesting to see you try Lifi internet. It could be the next big thing.

@AndreasSpiess Ай бұрын

It is on my list. But so far, a lot of talk but not a lot of actions :-(

@G0TT3RFUNK3N Ай бұрын

Halow advantage is versatility of working great indoors and outdoors without switching between 2.4 or 5.8 and the long range without bulky directional antenna just a regular omni stick like a ham radio. Hope manufacturers will focus on this 802.11ah/s use case scenarios and develop higher power add mimo for more thoroughput and more watt power for extended range like 6 to 7 miles. And make it man portable with battery like a tactical manet radio in a ruggedized tactical form factor.

@AndreasSpiess Ай бұрын

It will be hard for them to increase power as this is a legal issue, not a technical one. In non-licensed bands, regulators want to limit the range to not disturb others...

@newsogn5148 Ай бұрын

That’s great you can extend internet access, but node red runs locally can you use these to connect 2 devices on the same network like an esp32 and node red on opposite sides?

@AndreasSpiess Ай бұрын

These devices extend your network. You can connect whatever you desire if it does not consume high bandwith. So Node-Red is ok.

@pu5epx Ай бұрын

I see this as "Zigbee killer". Espressif could launch a chip that supports this.

@AndreasSpiess Ай бұрын

They serve a completely different purpose. I love Zigbee because it runs for more htan a year on one coin cell.

@MrWlcheck Ай бұрын

Nice video! Maybe you can try to make "reflection" parabola for HaLow to make it directional - i was using aluminium sheet with paper as holder some time ago for 2,4ghz

@AndreasSpiess Ай бұрын

It would be easy to create a directional antenna. But it is not allowed to change ISM devices. So I will not show it on KZfaq.

@MrWlcheck Ай бұрын

@@AndreasSpiess all right, this is what I didn't know. Thank you!

@rastikw Ай бұрын

@@AndreasSpiess As you as the importer are responsible for making sure it stays within the limits, you can do whatever you think will pass the regulations.

@GeoffHou Ай бұрын

It's a good point to verify it's within the European regulations. I still wonder however why the supplier of the RTK board of last week's video does not ship to the EU but does to Switzerland. Is there a import ban because of a component?

@tarakivu8861 Ай бұрын

I just know, if they havent setup tax stuff its not shown for shipping to EU countries.

@AndreasSpiess Ай бұрын

I do not think legal issues prevent the shipment of RTK receivers to the EU...

@McTroyd Ай бұрын

Dumb question, as I'm sure you checked, but could your 2.4 GHz result be because of interference? You mentioned 5.8 GHz is pretty busy in your area, but I know there are fewer effective channels at 2.4. In any case this is pretty cool. I could see it being a fun toy in field deployments. 👍

@AndreasSpiess Ай бұрын

Well possible that interference was the problem. I did not test it because I had a working solution on 5.8 GHz. 2.4 for sure is more crowded in most places.

@SCIENindustries Ай бұрын

in outdoor tests did you had clear line of sight? I need some solution to get internet connection to my solar station about 400m from home but there are trees between, so I'm not sure how good 5GHz is for that solution?

@AndreasSpiess Ай бұрын

I had a clear line of sight. Trees are not good for 5.8 GHz

@zrx2779 Ай бұрын

please make a test IN THE CITY - when the base will be located inside some building (office type) and you will go into another building which will be located for example 300 meters away + there will be 2-3 buildings in between because this tech has been designed not for the use inside of one building but for interconnect between different thanks!

@AndreasSpiess Ай бұрын

After my tests, you can imagine how this goes. Each case is different. So you have to do a test in your actual situation. Do not expect long distance and through walls. Ethernet or is more probable.

@-someone-. Ай бұрын

Bro thinks all the buildings in the city follow Andreas Spiess and grant him access to do internet tests😅... it would be good tho

@tarakivu8861 Ай бұрын

Lots of interference in a city too.. any as soon as more people use this, performance drops rapidly anyways.. too little bandwidth left. Might aswell bounce some 2.4 or 5.8GHz signals around the buildings and hope you get a multipath connection.

@zrx2779 Ай бұрын

@@AndreasSpiess thank you for your reply! but i (and i think all of your viewers) will be much happy to see a real life result rather than "results of our imagination" so PLEASE make such test 😉 and by "test" i mean being in real city such as Zurich (not a village-type small city) - you could place "base" antenna in some coworking office and then go to the street and show 3-5 types of result based on distance and location (for example 1 - you are just outside and 100 meters away, 2 - you are inside of another building, in some cafe, also 100m away, 3 - you are 300 meters away outside) because if we are trying to compare such tech with LORA - the lora works in the city on distances as 2-3 kilometers!! = 3000 meters! and if this tech will not be able to make a transfers even at distances of 100 meters - then it will have nothing similar with lora but just another hype for wifi

@wrekced Ай бұрын

I think the 2.4aGHz low performance might have something to do with water in the atmosphere absorbing more at that frequency.

@AndreasSpiess Ай бұрын

I am not sure, I never read about such a problem (I know of course why we use 2.4GHz for microwavers). Another cause coud be interferences.

@3xblahblah Ай бұрын

Might be interesting for motorcycle intercoms 🤔

@AndreasSpiess 28 күн бұрын

Maybe. Usually, we drive quite close so that BT works.

@skrywenko6596 Ай бұрын

curious what the difference between this and the normal NLOS wifi equipment that existed for decades such as the ubiquiti SuperRange9 from around 20 years ago.. I 've use these alot and setup many 10 mile connection as LOS and and +1mile as NLOS. though they fell out favour about 10 years ago due to their slower speeds and the faster 2.4 and 5 ghz wifi equipment

@AndreasSpiess Ай бұрын

I assumed the purpose was the same, but they used a different protocol.

@milensvobodnikov8484 Ай бұрын

Read about ack timeout and how it limits distance. Even with directional antennas, you can't go more that 300 meters with home AP.

@AndreasSpiess Ай бұрын

Maybe this was the problem on 2.4GHz. As you saw, at 5.8, more han 1 km was no issue.

@y__h Ай бұрын

2.4 ghz lower speed could be because of air humidity or interference from other wifi or devices around your neighborhood?

@AndreasSpiess Ай бұрын

Humidity is less problematic. It was an ok day. Maybe interferences. 2.4GHz is quite crowded. But it seemed, 5.8 too…

@georgepapa2950 Ай бұрын

@@AndreasSpiess Next time take your spectrum analyzer with you, also 2.4GHz has caused too much interference (wider transmission) compared to 5GHz wavelength shorter. I used 5.8GHz over a 20KM link in 2010 over the Congo River with a 32DB gain dish worked awesome those days with a real bandwidth of 60MBps

@timballam3675 Ай бұрын

Are these devices still governed by the 1% duty cycle ie 36s transmitting per hour?

@dblock8686 Ай бұрын

Very nice technology insights. How does Halow achieve such high data rate compared to LoRa while not exceeding the EU duty cycle limits?

@AndreasSpiess 28 күн бұрын

It uses a wider bandwidth and avoids strict duty cycle rules

@dblock8686 28 күн бұрын

@@AndreasSpiess how does it avoid duty cycle limits without being illegal

@conorstewart2214 Ай бұрын

It would be great if espressif made a module that used this.

@AndreasSpiess Ай бұрын

We will see...

@hypercube33 Ай бұрын

Would you be able to add the 2.4/5.8Ghz stuff to links? I think those sell themselves for a project I have and I'd like to support you!

@AndreasSpiess Ай бұрын

I added Amazon links. Thank you for your support!

@hypercube33 10 күн бұрын

@@AndreasSpiess Thanks! I'm likely going to invest in one or both of these for some testing and projects here shortly. This video is a deep dive talk about HaLow in case anyone wants to check that out. kzfaq.info/get/bejne/payGnZSDqrmdhIk.html I do agree, we need to push Wi-Fi lower frequency for PTP links where its possible due to line of sight concerns and other cool features this technology has.

@tonyh6309 Ай бұрын

Andreas, are you sure that end users are not allowed to change antenna on so-called short range radio (SRD) equipment? Is this a specific Swiss regulation or are you referring to equipment where the TX power cannot be reduced by the end user to compensate for different antenna gain? Obviously most users will not be in a position to measure the actual ERP of their equipment if they do change antenna and thus risk breaking the law. In practice it's likely very rare that the actual gain of an antenna significantly exceeds the specification - especially for cheap Aliexpress/Ebay antenna where the claimed gains can be way in excess of reality.

@AndreasSpiess Ай бұрын

AFAIK, all ISM devices have to be certified including antennas (and software). So if you change anything, it is no more certified. You decide if this matters for you ;-)

@tokenr7414 Ай бұрын

Outdoor 2.4GHz is heavily influeced by weather I think. We know that water molecule react to those frequencies...

@rkan2 Ай бұрын

Hence microwaves running on 2,4Ghz...

@AndreasSpiess Ай бұрын

Good point. However, during my tests, the weather was ok (no rain).

@gabest4 Ай бұрын

We have a wireless thermostat in the house that works on 868MHz. A traditional one, not IoT. If this extender continuously used all bandwidth, because for example I used it to connect security cameras in the garden, could they interfere? The thermostat gives out signals to turn on or off the heater every five minutes, what if the receiver cannot hear it?

@AndreasSpiess Ай бұрын

You would have to coordinate the frequencies. Both on the same frequency would not be a good idea…

@self_taught_stuff Ай бұрын

i really cant understand why companies dont put PoE option to their new devices.

@jorper2526 Ай бұрын

Because it's a cheap crappy Chinese device which is made inherently insecure by design.

@self_taught_stuff Ай бұрын

@@jorper2526 just bought 10 PoE injectors incorporated in the wire, calculated all in all for $3.2USD each, no changes needed to the design of the device at all, they are external on the wire and are even waterproof...

@AndreasSpiess Ай бұрын

PoE also needs a PoE enabled switch or router and not everybody has one...

@self_taught_stuff Ай бұрын

@@AndreasSpiess true that, but example the cheap cameras i buy have external PoE converter on the LAN cable and a 12V jack. i checked and those external PoE converters cost like $2USD

@leroymay8156 Ай бұрын

Interesting would be a outdoor connection trough an obsticle like a tree or a building. So no direct line of sight. In this case 5GHz should fail, but the low frequency HaLow could still work. That actually would be quite interesting for me, since my only HF connection for HAMNET is via NPR70, because there is a building between the Access Point on the Relais and my QTH. NPR70 wich works in the 70cm band works fine without line of sight, but it is extremely slow. (talking 64k dialup modem speeds...) Such a HaLow could be a compromise. Specially because the distance of my QTH and the Relais is only 780m. Would it be possible that you perform an additional test? thx es vy 73

@AndreasSpiess Ай бұрын

You can forecast the behavior of a link by adding my two tests: You know that it works for 1.5km and it does not like walls or buildings. So, on a short distance, it might pass walls. Trees are similar to walls. 5.8GHz does not like tham. In the US, it is for sure a competitor to NPR70.

@nrdgrrrl Ай бұрын

It looks like there's room for a few 18650s in the receiver.

@AndreasSpiess Ай бұрын

For sure. But it would interfere with the antennas...

@JoeRKsChannel Ай бұрын

What about duty cycle considerations? Surely a constant link like WiFi opposed to bursty LoRaWAN means these duty cycles will be ignored?

@AndreasSpiess Ай бұрын

I did not find duty cycle info in the 802.11ah documents

@BobHannent Ай бұрын

I've been using HaLow Wireless for a few years now to connect our holiday home to my father-in-law's house. It's on a mountain, about 600m P2P, but getting 5GHz working wasn't great, there are some trees that i cannot punch through. I found some "wireless IP CCTV extenders" and they've worked well. I get about 13Mbps out of it. The omnidirectional antenna didn't work well, so i have a TV aerial attached (best i could get rurally in a pinch).

@AndreasSpiess Ай бұрын

Indeed, 5.8GHz does not like trees ;-) And thank you for sharing your experience!

@siberx4 Ай бұрын

1:32 Note that this is a bit misleading. There's actually no direct frequency-dependent "path loss" that increases with distance; the "frequency" term in the usual free-space path loss equations comes from the fact that higher frequencies use smaller antennas for the same gain, so they're covering a smaller physical area and thus capturing less signal at a given distance. If your antennas are the same size at different frequencies (say, you're using dish antennas with different feedhorns) then this apparent effect disappears. The higher frequencies will get higher gain from the same dish, cancelling out the supposed "path loss" from that frequency. Consider laser communication up in the terahertz; it's not getting hundreds of dB of attenuation from distances of hundreds of kilometers, or space-based laser links would never be a real idea. You still have to deal with atmospheric absorption once you get up into the gigahertz range, but this is a separate effect that's nonlinear with frequency (various "windows" exist with lower losses, and it varies by moisture content in the air).

@AndreasSpiess Ай бұрын

AFAIK the path loss and the antenna gains are two different things and are not related other than they add up.

@siberx4 Ай бұрын

@@AndreasSpiess Sort of; the free-space path loss calculations assume an isotropic antenna (which makes it easy to add your antenna's actual gain to the path calculations), which is progressively smaller (and thus captures less signal area) the higher the frequency is. This is what causes the frequency dependent "path loss" component; if your dipole is only 30mm long (at say 5GHz), it's collecting a smaller percentage/fraction of the emitted radio energy when it's 1km away from the source compared to a 150mm antenna (at 1GHz) at the same distance. It makes the calculations straightforward, but it doesn't actually mean the higher frequencies are losing "more energy" somehow over distance (except by atmospheric effects for microwave+ frequencies, which are unrelated to the free-space path loss which assumes "free space"). The signal is emitted as a widening "bubble" of RF energy. The further you get, the more it spreads out, and the weaker the resulting signal. The bigger your receiving antenna, the more of that expanding bubble you capture at a given distance. A 0.5m parabolic dish has a gain of 12dB at 1GHz, but 26dB at 5GHz. Considering this on both the transmitting and receiving ends, this compensates for (corrects for) the increased frequency-dependent "path loss" component because with a parabolic dish, both frequencies get the same swept area (not true for simpler antenna designs like dipoles). Presented another way, if you had a 30dBi antenna for a 1GHz link and a 30dBi antenna for a 5GHz link, the 1GHz link would perform better - but the antenna would also be huge (4m dish vs 0.8m dish). Same size dish = same performance (assuming equal transmit powers and receiver sensitivity).

@theknivjocke Ай бұрын

I think the issue with 2.4 GHz is lots of interference, microwave ovens and other legacy devices. 5.8 is quieter, and other traffic is attenuated more due to the wavelength.

@AndreasSpiess Ай бұрын

I agree.

@retokaderli9010 Ай бұрын

How about the 433MHz range? Can I extrapolate? even better penetration indoors, but lower throughput,, ... so, just for some sensor data, but not for home networking bridges.

@G_C340 Ай бұрын

You are then into amateur bands and far more possibility of disruption from other signals.

@AndreasSpiess Ай бұрын

Every frequency band has its properties. Below 30MHz, signals from my radio station can reach New Zealand ;-)

@neildotwilliams Ай бұрын

They may not be able to sell these in the UK due to a new law. Default passwords are bad. Product Security and Telecommunications Act 2022 (PSTI). Fines incoming 😶

@tarakivu8861 Ай бұрын

This might affect normal sellers, but the Aliexpress seller couldnt care less :D

@neildotwilliams Ай бұрын

@@tarakivu8861 Yes very true.

@major__kong Ай бұрын

The decrease in range with frequency is not directly related to frequency rather the use of resonant (dipole) antennas that get shorter with increasing frequency. Use a parabolic dish and range will go up. But then it isn't omni anymore :-)

@AndreasSpiess Ай бұрын

I have a 2.4GHz dish on my roof where I can talk to a geo.stationary satellite 36'000km away ;-)

@saxpert Ай бұрын

Carriers noticed immediately when the noise figures changed due to other transmitters. They will inform the government (BNetzA in Germany) and they will find the jammer quickly.

@AndreasSpiess Ай бұрын

Thank you for the detailed info!

@bjornroesbeke Ай бұрын

6:36 14dBm, but how does the gain of the antenna influence this value?

@tarakivu8861 Ай бұрын

Its probably omnidirectional antenna, so the apparent transmitted power will be slightly higher (2-4dB)

@AndreasSpiess Ай бұрын

I would not count on gain from such an omnidirectional antenna.

@berlinberlin4246 Ай бұрын

Any new long range equipment without WPA3 support is a no go, As WPA2 and spacaly WPA is to broken to can be safely used 😢

@abavariannormiepleb9470 Ай бұрын

With all your experience can you recommend a manufacturer for a building-to-building Wi-Fi bridge with 60 GHz and Wi-Fi 6 ax fallback mode for a distance of 100 m? I’m looking for a manufacturer with proper long-term standalone firmware update support, cloud subscription requirements for proper functionality is a disqualifying issue for me.

@AndreasSpiess Ай бұрын

I thought, I made a video about the 60/5GHz link ? It worked across 1.5km...

@TymexComputing Ай бұрын

Oh - is there anybody here familiar with the terminology/keyword to search - regarding the Wireless Opticall Point to Point transmission lines? Heard about it 15 years ago - look like a camera, pointed to another one, IR transmission modulated. should be able to receive ~100-500 mbps at 100-500 meters :). Wondering whats the modulation used in it, whats are the details - and any patent numbers to read would be appreciated :)

@AndreasSpiess Ай бұрын

I cannot help :-(

@dbird29 Ай бұрын

More noise floor on 2.4 GHz?

@AndreasSpiess Ай бұрын

Maybe (with all my ESP sensors in my home ;-)

@dfgaJK Ай бұрын

16:15 so is a HAM in the US actually allowed to use HaLow at 1.5kW?

@AndreasSpiess Ай бұрын

I am not sure. Just found an ARRL band plan where it was mentioned. Maybe a US HAM knows for sure. Anyway. 10 or 100 W would cos sure already move the needle…

@paaao Ай бұрын

2.4 has a much higher noise floor due to it's ability to leak out of all the surrounding buildings and homes

@AndreasSpiess Ай бұрын

Well possible (with all the Espressif stuff around my house)...

@ws_stelzi79 Ай бұрын

Well considering Switzerland has basically more bunkers for people than normal houses it seems very difficult even for Swiss people to detect when they are living in bunkers with reinforced concrete. Also bunkers are not "optimized" for WiFi. 😉😇

@AndreasSpiess Ай бұрын

The difference is the windows ;-)