I’m a massive networking nerd. I have literal stacks of old networking hardware, probably enough to connect a small town. It’s almost all used and some is damaged and I love the shit out of every scrap circuit board with those glorious ports.
I usually end up ranting about home networking on Lemmy, and the networking subreddits are generally the only reason I go back to that site every now and again.
I’ve become a wireless expert, and I regularly flex that knowledge at work. It always amazes me how bad some people’s wifi is and they just accept it, like, it do be like that sometimes… But it doesn’t have to be like that.
Because of this I often find myself ranting about what to do, or not do, when it comes to home networking projects. I always feel like this falls on deaf ears because I end up repeating the same or similar rants regularly.
You know HAM radio? It’s kinda interesting, because people can use it to talk to each other and it technically doesn’t require any infrastructure. But there are also repeaters in cities that can increase the range.
I’ve been wondering why people don’t try to do something similar with WiFi? Some kind of city-wide WiFi network with repeaters. It’s probably difficult and I’m not sure if it would have any practical use. But the advantage over HAM radio is that it’s encrypted and doesn’t require a license. I imagine that people could use it to chat with each other and share stuff without having to rely on social media or the internet.
I’m a qualified amateur operator. I can operate on any ham bands up to 190W EIRP if I recall my countries regulations correctly.
The issue with doing something like wifi on ham bands technical issues finding radio chips capable of signalling at such a high rate of speed, on bands that are able to be used by hams. There’s also the requirement for hams to identify themselves on air, and the general use of AM/FM and derivative technology on ham bands and general resistance to the OFDM used as the main signal encoding for WiFi. So finding an OFDM capable radio transmitter/receiver for use in… say, the 2m band (144Mhz … ish) is basically impossible, and there’s no way to identify. You would have to build a new protocol and standard from the ground up and use very modified or rare/expensive radio chips, and likely build the drivers/firmware for it entirely yourself. People with the required hardware, software, baseband, radio, and firmware experience that are hams who want a product like WiFi for ham radio channels is extraordinarily rare.
As for city-wide WiFi/mesh networks: it has been attempted, and has seen some limited success, but doesn’t scale well with the usual protocols. Routing protocols like BGP, OSPF and IS-IS are meant for much larger IP blocks being routed between interfaces. A wireless mesh system would use a single interface (one radio) for both send and receive, which most protocols don’t support, and each “hop” or station on the mesh would only be advertising a single IP (or an extremely small set of IPs) per participating node.
Most routing protocols assume that every node on an interface can talk to every other node on the same interface and thus there’s no need to repeat or relay messages from an interface to the same interface.
There’s also no standards that allow wifi to use multiple channels/frequencies for tx/rx, eg, send on 5.45 GHz, and receive on 5.65Ghz. it simply isn’t something that any WiFi chip is capable of. So full duplex isn’t possible right now.
The common wifi frequencies are also extremely power limited and on bands that are prone to interruption. In the wild, there’s plenty of things that can disrupt 2.4Ghz and 5/6Ghz transmissions. With the power limits, to go any significant distance, you need directional antennas that limit free space path loss so the signal travels further. In the case of wireless internet service providers (WISP, not to be confused with the mobile carriers), they generally use panel or dish antennas to extend the range.
For power output, at the high end, some bands allow for upwards of 5W of directional power, or 1W of omnidirectional power (in EIRP). On the low end, handheld ham radio units start at 5W of power, and can usually attenuate their transmitter to 1W or lower as an option. Household WiFi is usually around 0.1W of power per radio. Even cranking that up to the maximum legally allowed wattage won’t result in covering more than a few blocks of a city with a fairly poor signal overall; that signal is going to be fairly easily blocked, absorbed, reflected, or otherwise attenuated by just about everything, including, but not limited to the structure of your house.
Meanwhile, standing in my home with a 5W handheld transceiver operating on 70cm (440Mhz), I can hit a repeater that’s something like 10 miles away with a nearly perfect signal over FM, without assistance. OFDM signals would likely be scrambled beyond recognition at 2.4 or 5 GHz across that much of a gap, or even one that’s 1/10th as far, with only 1W of legal power, without using some kind of directional antenna or antenna array.
Don’t get me wrong, well configured wireless can go so far that you have to account for the curvature of the earth, but they’re always very very directional, using dish antennas or similar.
Don’t get me wrong, the ideas are great, but the challenges faced are enormous. It can quickly turn into a lifelong project to get something functional, and even then, there’s no guarantee that it will ever catch on as a product. The limitation for ham operators regarding encryption is problematic when it comes to data communication as well, since just about everything that’s data-driven on the internet implements SSL. Computers and systems expect encryption all over the place and bluntly, those messages cannot be sent over ham bands. There’s nuance to that regulation, at least in my country, but I won’t get into the fine print here.
Even so, there are some crude digital modes used by ham operators which are normally voice encoding or plain text encoding. Uses are limited on purpose. If you’re interested in longer distance emergency communications you could look into LoRA, which is relatively new.
There’s a lot more to say on this, but bluntly, I’ve said enough. It’s all interconnected, and I love it, but I’m just ranting now.
Thanks for a very detailed explanation! So it seems that this is almost impossible. Except for maybe a small part of a city.
The lack of encryption, privacy and anonymity in HAM radio would be an issue for me. Just like in mobile phones. But since you can use end-to-end encryption over WiFi (and some weaker, less useful encryption is used in mobile phones), maybe there are exceptions. I’ve heard of LoRA, but I wouldn’t want to use it, unless I’m allowed to encrypt the messages. It also seems that the message length is very limited, so using something like PGP might not always be possible.
My favorite so far was a factory that put a giant metal shipping container in the middle of the shop between the AP and where they were using the wifi and complained that they couldn’t understand why that part of the shop got no signal.
I’m a massive networking nerd. I have literal stacks of old networking hardware, probably enough to connect a small town. It’s almost all used and some is damaged and I love the shit out of every scrap circuit board with those glorious ports.
I usually end up ranting about home networking on Lemmy, and the networking subreddits are generally the only reason I go back to that site every now and again.
I’ve become a wireless expert, and I regularly flex that knowledge at work. It always amazes me how bad some people’s wifi is and they just accept it, like, it do be like that sometimes… But it doesn’t have to be like that.
Because of this I often find myself ranting about what to do, or not do, when it comes to home networking projects. I always feel like this falls on deaf ears because I end up repeating the same or similar rants regularly.
You know HAM radio? It’s kinda interesting, because people can use it to talk to each other and it technically doesn’t require any infrastructure. But there are also repeaters in cities that can increase the range.
I’ve been wondering why people don’t try to do something similar with WiFi? Some kind of city-wide WiFi network with repeaters. It’s probably difficult and I’m not sure if it would have any practical use. But the advantage over HAM radio is that it’s encrypted and doesn’t require a license. I imagine that people could use it to chat with each other and share stuff without having to rely on social media or the internet.
I’m a qualified amateur operator. I can operate on any ham bands up to 190W EIRP if I recall my countries regulations correctly.
The issue with doing something like wifi on ham bands technical issues finding radio chips capable of signalling at such a high rate of speed, on bands that are able to be used by hams. There’s also the requirement for hams to identify themselves on air, and the general use of AM/FM and derivative technology on ham bands and general resistance to the OFDM used as the main signal encoding for WiFi. So finding an OFDM capable radio transmitter/receiver for use in… say, the 2m band (144Mhz … ish) is basically impossible, and there’s no way to identify. You would have to build a new protocol and standard from the ground up and use very modified or rare/expensive radio chips, and likely build the drivers/firmware for it entirely yourself. People with the required hardware, software, baseband, radio, and firmware experience that are hams who want a product like WiFi for ham radio channels is extraordinarily rare.
As for city-wide WiFi/mesh networks: it has been attempted, and has seen some limited success, but doesn’t scale well with the usual protocols. Routing protocols like BGP, OSPF and IS-IS are meant for much larger IP blocks being routed between interfaces. A wireless mesh system would use a single interface (one radio) for both send and receive, which most protocols don’t support, and each “hop” or station on the mesh would only be advertising a single IP (or an extremely small set of IPs) per participating node.
Most routing protocols assume that every node on an interface can talk to every other node on the same interface and thus there’s no need to repeat or relay messages from an interface to the same interface.
There’s also no standards that allow wifi to use multiple channels/frequencies for tx/rx, eg, send on 5.45 GHz, and receive on 5.65Ghz. it simply isn’t something that any WiFi chip is capable of. So full duplex isn’t possible right now.
The common wifi frequencies are also extremely power limited and on bands that are prone to interruption. In the wild, there’s plenty of things that can disrupt 2.4Ghz and 5/6Ghz transmissions. With the power limits, to go any significant distance, you need directional antennas that limit free space path loss so the signal travels further. In the case of wireless internet service providers (WISP, not to be confused with the mobile carriers), they generally use panel or dish antennas to extend the range. For power output, at the high end, some bands allow for upwards of 5W of directional power, or 1W of omnidirectional power (in EIRP). On the low end, handheld ham radio units start at 5W of power, and can usually attenuate their transmitter to 1W or lower as an option. Household WiFi is usually around 0.1W of power per radio. Even cranking that up to the maximum legally allowed wattage won’t result in covering more than a few blocks of a city with a fairly poor signal overall; that signal is going to be fairly easily blocked, absorbed, reflected, or otherwise attenuated by just about everything, including, but not limited to the structure of your house.
Meanwhile, standing in my home with a 5W handheld transceiver operating on 70cm (440Mhz), I can hit a repeater that’s something like 10 miles away with a nearly perfect signal over FM, without assistance. OFDM signals would likely be scrambled beyond recognition at 2.4 or 5 GHz across that much of a gap, or even one that’s 1/10th as far, with only 1W of legal power, without using some kind of directional antenna or antenna array.
Don’t get me wrong, well configured wireless can go so far that you have to account for the curvature of the earth, but they’re always very very directional, using dish antennas or similar.
Don’t get me wrong, the ideas are great, but the challenges faced are enormous. It can quickly turn into a lifelong project to get something functional, and even then, there’s no guarantee that it will ever catch on as a product. The limitation for ham operators regarding encryption is problematic when it comes to data communication as well, since just about everything that’s data-driven on the internet implements SSL. Computers and systems expect encryption all over the place and bluntly, those messages cannot be sent over ham bands. There’s nuance to that regulation, at least in my country, but I won’t get into the fine print here.
Even so, there are some crude digital modes used by ham operators which are normally voice encoding or plain text encoding. Uses are limited on purpose. If you’re interested in longer distance emergency communications you could look into LoRA, which is relatively new.
There’s a lot more to say on this, but bluntly, I’ve said enough. It’s all interconnected, and I love it, but I’m just ranting now.
Thanks for a very detailed explanation! So it seems that this is almost impossible. Except for maybe a small part of a city.
The lack of encryption, privacy and anonymity in HAM radio would be an issue for me. Just like in mobile phones. But since you can use end-to-end encryption over WiFi (and some weaker, less useful encryption is used in mobile phones), maybe there are exceptions. I’ve heard of LoRA, but I wouldn’t want to use it, unless I’m allowed to encrypt the messages. It also seems that the message length is very limited, so using something like PGP might not always be possible.
You ever seen a Unify AP just chilling on the floor under a desk in a commercial setting while people complain the WiFi is always slow?
Pain.
Or when they put it inside or behind a metal file cabinet and wonder why the wifi doesn’t work unless you’re right next to the file cabinet
My favorite so far was a factory that put a giant metal shipping container in the middle of the shop between the AP and where they were using the wifi and complained that they couldn’t understand why that part of the shop got no signal.