As usual, I love read what traditional and “tech” press can write about LPWAn technologies. That’s always a pleasure to read such amount of bullshit condensed is a maximum of 10-15 lines. They are directly copied and paste from a press release. The Author added value is a custom and totally wrong clickbait title.
As part of many, the one making me starting this post is from “Le monde informatique”. It is supposed to be a serous source. The title was “sidewalk a threat for LoRa and Sigfox on IoT”.
Just to let you know guys… Sidewalk is just a new Zigbee !
So let see what is really Amazon sidewalk, what are the strength and weakness of this solution
What is Amazon Sidewalk
It is a low power, long range RF solution
Based on the Amazon announcement, Sidewalk is a RF protocol based on ISM bands (900 Mhz in the USA). It is a low power RF solution able to run IoT devices like door bell, pet trackers…
This solution seems to cover a range of 500 meters and can’t be considered as a Wide-Area network like LoRaWan, Sigfox and Nb-Iot are. This can be more compared as a Zigbee solution. The closed brother technology should be DASH7 as this one supports meshed networks.
That said, the 500 meters can also be the consequence of using gateways installed indoor, like in a Amazon echo systems. The use of such type of gateway, like for LoRaWan, could highly restrict the range of the solution, even if the technical solution can propose a larger range.
It’s interesting to see that Ring company (the one making Fetch device on the photo) has recently certified a spotlight product based on LoRa technology for communication.
Amazon sidewalk is a mesh network. It means devices are communicating from device to device to reach a given destination. It can be another device in the network but we can assume it will mostly try to join a gateway to reach Internet and join a server then end-user mobile application.
The problem of meshed network in my point of view is it creates complexity.
- The main of all is related to the energy consumption (because a device in a mesh network needs to listen the RF spectrum and this has a large energy cost). For this reason I assume we are going to see different class of devices: mesh nodes (listening and repeating) and mesh client (just talking)
- The second one is not applicable in USA where the ISM band can be use 100% of the time until you transmit on different frequencies. If you need to listen and repeat a message you need to be able to talk frequently with no delay. But in Europe the ISM bands requires 1% duty cycle in most of the bands. You can eventually target 10% but the only small band. The frequency where you can communicate 100% of the time have a really reduced transmission power and are too small to make a network on it without making lot of collisions. Most of the meshed network prefer 2.4GHz were there is no duty cycles.
- The last one is the complexity to managed such meshed network, including addressing. The consequence is a fat firmware and more expensive MCU to run it. It is also a larger surface for attack and a more difficult solution to protect.
You see I’m not a big fan of the meshed network but for sure they can have some really interesting advantages. Here the example is to have a dog or a cat moving around and even if it is out of your home network coverage you can continue to track it thanks to neighbor networks.
But this makes sense only if you are able to jump from a point to another point. I mean if the coverage is large enough.
City wide network
The 700 nodes Amazon proof of concept
The example given by Amazon is a test in Los Angeles with 700 nodes deployed in the city and that “covered much of Los Angeles Basin”. Ok, so that’s not all Los Angeles basin. Are these nodes the Ring spotlight product is a good question …
So just to let you know, to cover Los Angeles Basin with LoRaWan, you may need 20-30 antennas. With Sigfox about 10. Just to compare.
What is interesting with Amazon is that you can expect to have about 700 people or more deploying Amazon Sidewalk nodes in a such city. Thanks to the marketing capacity they can make it and faster than anyone. Just imagine all the Amazon Echo have a such tech in it…
The choice of the tested node is interesting: they deployed ring lighting product. This sounds an outdoor product with a large number of batteries in it. In terms of coverage, having node outdoor vs indoor makes a big difference. As a comparison LoRaWan indoor gateway covers about 500m around when an outdoor one covers about 10km.
Would it work the same world-wide?
The last point to take care is the transmission power. ISM bands in North America accepts +20dB transmission power. It means that in Europe with only +14dB, the given 500m may be reduced to 200-300m or the transmission speed could be affected. 6db difference is 4 times less power.
Can we consider a mesh as a public network equivalent?
We can’t really compare Amazon Sidewalk with LoRaWan or Sigfox: this is not a managed network with central gateways and network operator. The coverage will depends on the customers. In a first time you will only cover you garden and only have application to use on your own. Later you may have a larger network for city wide usage. This is like the telecoms WiFi networks based on WiFi sharing between operator’s consumer. It exists and sometime you are able to use them.
You have no chance to cover country-side, mountains, roads where the people density is low. The application are really reduced to cities. As an example: your dog tracker may work if your animal leaves the house to visit your city neighborhood. But if it leaves when you make your Sunday walk in a forest you will never get it back.
Amazon Sidewalk challenges to eat the world
Where the regulation make it complex
Regulation are different in the different zones. For Sigfox and LoRaWan it has been a long journey to find the technical solutions to make the devices and gateways working with these regulations. And most of the devices still work only in a given zone. Technologies like Sigfox Monarch has to be deployed to make device able to take the right configuration at the right place.
It will be the same for Amazon Sidewalk. Managing the different frequencies, duty cycle restriction, and power restriction will really slow down the technology deployment. North America will be available next year but what about the other zones?
The choice of 900Mhz ISM band (902-928MHz) for North America is a good choice for long range RF solution. It means a use of 868Mhz in Europe and frequency around these all over the world. But compared to 2.4GHz or 433MHz, there is no world-wide standards on that frequency. Dash7 has chosen 433MHz for this reason: being able to cover the world with one frequency and compatible regulations.
And it makes more difficult to make devices
The consequence is a hardware segmentation: the device providers will not have devices compatible for all zones at start. For a world-wide distributor like Amazon it means a certain complexity to manage. Nothing they can’t. As a consequence, for the consumer, it’s a bit discouraging to see product existing in USA, you can’t buy and use in Europe or Asia.
I agree the North American market is large enough to make it a success. But, Amazon communication world-wide is now generating a wait all-around the world.
How bad can be Amazon Sidewalk
A side effect of a such technology is the RF noise it will generate. Amazon Sidewalk may be a source of many long range, high power communications. This will be due to the complexity to communicate with an indoor gateway. This indoor gateway (like an Amazon echo) is located in a noisy environment. As a consequence the solution may generated a lot of outdoor noise on the small ISM band.
Depends on the modulation used it can create collision with networks like LoRaWan or Sigfox.
This risk is large as Amazon have a string marketing capacity to promote its technology. It can be a large success. We will see how it uses the RF spectrum. It can be respectful with technology like Ultra-Narrow band or at the opposite wide-band with a larger impact. The choice of the sub-band will be also something important to notice.
I’m curious to get details on the details of this technology to correctly highlight this point. I just hope it will not base on the Ring Spotlight product: this one has been certified for LoRa 500KHz, that’s really fat!
Is Amazon Sidewalk a threat for LPWAN?
Imagine Amazon Sidewalk based on LoRa modulation, it would be a serious technology push for Semtech. This company is the only one owning the technology and only radio-chip provider for it. So it could be a big jump until the radio-frequency become something totally not usable for both Amazon Sidewalk and LoRaWan due to collisions.
The potential choice of LoRa could be a great idea as the ecosystem is large on this technology.
By-the-way, LoRa is not LoRaWan. LoRa is a point to point radio technology, like Ethernet for cables. LoRaWan is a network on top of LoRa, like IP is on top of Ethernet.
Amazon sidewalk is not a network, it is not covering countries, that’s more to be compared with Zigbee even if potentially based on LoRa. The use-cases are totally different. As an example, Amazon Sidewalk does not work with supply-chain, extended mobility, tracking, water meter reading… use-cases. Some of the mostly covered by LPWAn. The domain of application for Amazon Sidewalk is smart-home and consumer gadgets where LoRa is but not really the LPWAn (LoRaWan and Sigfox)