5 ms·

Whoof. A pretty damning overview of one of the few intriguing "Web3" use cases. I've always thought blockchain subsidized mesh networks were a great idea - peo

Whoof. A pretty damning overview of one of the few intriguing "Web3" use cases. I've always thought blockchain subsidized mesh networks were a great idea - people provide access via a router, and are reimbursed based on how many other people access the network via their node. Seems ideal, and could quickly and easily scale across an urban area. Is there some inherent limitation in the underlying technology that prevents this type of distributed system?

amluto 4y ago
Yes, two of them. First, RF bandwidth is a shared resource. One may credibly complain about the FCC’s allocation of bandwidth, but the big cellular networks work well because they are centrally managed. If every tall building had ten antennas operated by competing token miners, then they would probably have incentives to deliberate interfere with each other, and they would have little incentive to cooperate. The overall results would be much worse than what we have today. 2. Decentralized blockchains are inefficient. Helium would work every bit as well as it does today (hah) if the payments were centrally managed. Of course, it would be more to regulators and law enforcement, which might be a good thing.
- blacksmith_tb 4y ago
  I don't think the primary issues are technical (LoRA is packet-based, so there'd need to more than 10 antennas near each other to cause headaches), they're financial. Even if we grant the basic premise that it would be useful to have LoRA coverage all over the place, the operator of the network needs to actually have customers who will pay to use it, and pay enough to cover then paying the people hosting nodes. I think you're right that Helium's blockchain/crypto angle is more about avoiding oversight / taxes / liability than it is any sort of fundamental requirement of getting lots of people to sign up to join building out coverage.
jonathan-adly 4y ago
No - it is a great idea and will eventually work. A massive demand for this service doesn't really exist yet - so Helium is struggling. IOT devices everywhere are not here yet. But I don't see anything wrong with the underlying technology and the subsidy model. The only hiccup is the expected legal fight from big telco!
- rcoder 4y ago
  I'm working on a different LoRa-related startup, and I can say with pretty high confidence that most people hear "mesh network" and jump directly to, "so I can check Facebook from anywhere?" (Replace Facebook with Slack, Zoom, YouTube, or any other bandwidth-hungry, "use it all day" app you can think of.) And yet, the theoretical (not even real-world) bandwidth for LoRa is in the hundreds of bytes per second; perhaps whole _kilobytes_ if you're giving up most of the range and power savings that LoRa offers for traditional IoT. On the "small devices exchanging small amounts of data" front, LoRa, particularly for large-scale deployments like the ones Helium resells access to, faces a lot of competition from Thread, Matter, and other "home-scale" mesh options. Until Apple or Samsung puts a LoRa radio in their handsets, it's going to be a niche product. (Unlikely because Semtech owns all the patents on LoRa. One of them might _buy_ Semtech, but I imagine they'd shell out $2 per handset approximately _never_.)
  - digdugdirk 4y ago
    Do you have any good starting points for reading up on the differences between LoRa and traditional WiFi? Both technical and practical.
wmf 4y ago
Since Internet access has converged on an unlimited usage pricing model this will end up like Spotify: customers pay a flat rate which gets sliced extremely thin and providers get paid very little.
- LordDragonfang 4y ago
  I'm not sure it's correct to say that internet access has "converged" on unlimited usage. Most cellular plans have data caps (even "unlimited" plans have caps that they throttle after) and even hardwired internet plans are starting to have data caps by default now.
munificent 4y ago
As I understand it, the idea is that people are essentially subletting the bandwidth they pay for from their ISP. ISPs famously overprovision users. While your ISP plan may nominally guarantee some level of bandwidth, the network physically can't support all users using that much bandwidth at the same time. It's like gyms where if everyone who had a membership actually showed up, there wouldn't be enough gear to go around. The gym's business model completely relies on the fact that most members don't use the maximum of what their plan allows. ISPs don't want you subletting their bandwidth because that makes it easier to consume all of what your plan allows. They absolutely can't handle a signficant fraction of their users doing that. So, if a subsidized mesh network were to ever become popular enough to be a viable business, ISPs would simply forbid users from participating in it. It's just a fundamentally broken business model. It would be like sharing a gym membership across multiple people. Makes perfect sense for the member. "Hey, I only go one day a week anyway, so I can let 6 other people share it on the other days!" But the gym itself straight up could not sustain that at their current membership prices. Their only options are: 1. Forbid it (which is what they already do, since membership is associated with a specific person). 2. Go out of business. 3. Jack up membership prices until they can sustain all members fully utilizing their membership. But that would drive away members who didn't share memberships and make it so that your fraction of a shared membership is about as expensive as a single one used to be. Fundamentally, you can't pull infrastructure out of thin air. Pretending you can be reselling it won't work.
- sokoloff 4y ago
  Sharing a portion of my 300Mbps down/20 Mbps up bandwidth for a LoRaWAN network that passes dozens to low hundreds of bps per device is like fretting over people breathing the air in the gym parking lot.
  - munificent 4y ago
    If the device is only passing a few hundred bits per second, that also implies that it is doing almost nothing of value as well. Here's maybe another way to think about it: What possible viable business model is there for consumer ISP customers to build an ad hoc wireless network on top of the ISP's infrastructure that wouldn't immediately get killed by ISPs entirely reasonably cutting out the middle man?
Ensorceled 4y ago
> Is there some inherent limitation in the underlying technology that prevents this type of distributed system? The ToS of the ISP you plugged your "web3 router" into prevent you from reselling your bandwidth.
nradov 4y ago
Regardless of the payment model, construction of reliable, large scale wireless mesh networks remains an unsolved problem. Scaling those is not easy due to nodes going up and down, terminal devices moving around, intermittent RF interference, etc. The One Laptop Per Child project famously tried to use such a network in order to extend Internet access into areas lacking communications infrastructure, but it never really worked. This remains an area of active research and I'm skeptical whether the field has matured enough for commercialization.

amluto 4y ago

Yes, two of them. First, RF bandwidth is a shared resource. One may credibly complain about the FCC’s allocation of bandwidth, but the big cellular networks work well because they are centrally managed. If every tall building had ten antennas operated by competing token miners, then they would probably have incentives to deliberate interfere with each other, and they would have little incentive to cooperate. The overall results would be much worse than what we have today. 2. Decentralized blockchains are inefficient. Helium would work every bit as well as it does today (hah) if the payments were centrally managed. Of course, it would be more to regulators and law enforcement, which might be a good thing.

blacksmith_tb 4y ago
I don't think the primary issues are technical (LoRA is packet-based, so there'd need to more than 10 antennas near each other to cause headaches), they're financial. Even if we grant the basic premise that it would be useful to have LoRA coverage all over the place, the operator of the network needs to actually have customers who will pay to use it, and pay enough to cover then paying the people hosting nodes. I think you're right that Helium's blockchain/crypto angle is more about avoiding oversight / taxes / liability than it is any sort of fundamental requirement of getting lots of people to sign up to join building out coverage.

jonathan-adly 4y ago

No - it is a great idea and will eventually work. A massive demand for this service doesn't really exist yet - so Helium is struggling. IOT devices everywhere are not here yet. But I don't see anything wrong with the underlying technology and the subsidy model. The only hiccup is the expected legal fight from big telco!

rcoder 4y ago
I'm working on a different LoRa-related startup, and I can say with pretty high confidence that most people hear "mesh network" and jump directly to, "so I can check Facebook from anywhere?" (Replace Facebook with Slack, Zoom, YouTube, or any other bandwidth-hungry, "use it all day" app you can think of.) And yet, the theoretical (not even real-world) bandwidth for LoRa is in the hundreds of bytes per second; perhaps whole _kilobytes_ if you're giving up most of the range and power savings that LoRa offers for traditional IoT. On the "small devices exchanging small amounts of data" front, LoRa, particularly for large-scale deployments like the ones Helium resells access to, faces a lot of competition from Thread, Matter, and other "home-scale" mesh options. Until Apple or Samsung puts a LoRa radio in their handsets, it's going to be a niche product. (Unlikely because Semtech owns all the patents on LoRa. One of them might _buy_ Semtech, but I imagine they'd shell out $2 per handset approximately _never_.)
- digdugdirk 4y ago
  Do you have any good starting points for reading up on the differences between LoRa and traditional WiFi? Both technical and practical.

wmf 4y ago

Since Internet access has converged on an unlimited usage pricing model this will end up like Spotify: customers pay a flat rate which gets sliced extremely thin and providers get paid very little.

LordDragonfang 4y ago
I'm not sure it's correct to say that internet access has "converged" on unlimited usage. Most cellular plans have data caps (even "unlimited" plans have caps that they throttle after) and even hardwired internet plans are starting to have data caps by default now.

munificent 4y ago

As I understand it, the idea is that people are essentially subletting the bandwidth they pay for from their ISP. ISPs famously overprovision users. While your ISP plan may nominally guarantee some level of bandwidth, the network physically can't support all users using that much bandwidth at the same time. It's like gyms where if everyone who had a membership actually showed up, there wouldn't be enough gear to go around. The gym's business model completely relies on the fact that most members don't use the maximum of what their plan allows. ISPs don't want you subletting their bandwidth because that makes it easier to consume all of what your plan allows. They absolutely can't handle a signficant fraction of their users doing that. So, if a subsidized mesh network were to ever become popular enough to be a viable business, ISPs would simply forbid users from participating in it. It's just a fundamentally broken business model. It would be like sharing a gym membership across multiple people. Makes perfect sense for the member. "Hey, I only go one day a week anyway, so I can let 6 other people share it on the other days!" But the gym itself straight up could not sustain that at their current membership prices. Their only options are: 1. Forbid it (which is what they already do, since membership is associated with a specific person). 2. Go out of business. 3. Jack up membership prices until they can sustain all members fully utilizing their membership. But that would drive away members who didn't share memberships and make it so that your fraction of a shared membership is about as expensive as a single one used to be. Fundamentally, you can't pull infrastructure out of thin air. Pretending you can be reselling it won't work.

sokoloff 4y ago
Sharing a portion of my 300Mbps down/20 Mbps up bandwidth for a LoRaWAN network that passes dozens to low hundreds of bps per device is like fretting over people breathing the air in the gym parking lot.
- munificent 4y ago
  If the device is only passing a few hundred bits per second, that also implies that it is doing almost nothing of value as well. Here's maybe another way to think about it: What possible viable business model is there for consumer ISP customers to build an ad hoc wireless network on top of the ISP's infrastructure that wouldn't immediately get killed by ISPs entirely reasonably cutting out the middle man?

Ensorceled 4y ago

> Is there some inherent limitation in the underlying technology that prevents this type of distributed system? The ToS of the ISP you plugged your "web3 router" into prevent you from reselling your bandwidth.

nradov 4y ago

Regardless of the payment model, construction of reliable, large scale wireless mesh networks remains an unsolved problem. Scaling those is not easy due to nodes going up and down, terminal devices moving around, intermittent RF interference, etc. The One Laptop Per Child project famously tried to use such a network in order to extend Internet access into areas lacking communications infrastructure, but it never really worked. This remains an area of active research and I'm skeptical whether the field has matured enough for commercialization.