Good, I think? Fiber is technically superior in almost every way for ISPs. If incentives are aligning to build wider fiber networks then we all benefit.
I think 5G will be the death of the independent WISPs that don't adapt.
I have the T-Mobile 5G gateway as my sole home internet connection. I get a rock solid 300/100 with 30ms latency for $50/month. That approaches DOCSIS 3.1 performance, let alone WISP performance. Not sure how they will be able to compete.
This may be true for the majority of people and random web tasks. But performance is more than just thoroughput. The random round trip time and lack of ipv4 from mobile telcos means if you want to participate in the internet as a peer rather than just a consumer you need a physical transmission line.
>The random round trip time and lack of ipv4 from mobile telcos means if you want to participate in the internet as a peer rather than just a consumer you need a physical transmission line.
I run a weather station on T-Mobile. Yes it's ipv6-only inbound but a Cloudflare free plan is enough to provide legacy (ipv4) access to the station. While I agree that running a media server wouldn't be possible from this location, I disagree that it wouldn't work closer to a tower. Upload speeds there are over double what I can buy on Comcast business's most expensive plan.
You're right. 99.99% of users don't use the internet. They're just web consumers. As for hosting for people that actually use the internet, when you use a third party you have no expectation of privacy. Hosting from your own machines from your own home is always the best, easiest, cheapest, option.
That said, hosting from home and syndicating to a VPS is a great idea.
I always thought the use case for WISPs was for places that have at best the fiber line built for "a cell tower" but are otherwise cut off from more solid ISPs who want home internet or for places that are best have shitty ADSL. That of course means people in distant nowhere.
I never thought of WISPs existing elsewhere as it doesn't make nearly as much sense if you have access to "real" broadband. I suppose this niche was wider at one point?
20 years ago, wireless was the only affordable way for us to serve residential customers faster than ISDN or in some cases dialup in a lot of small towns. We could get FR or ATM to a business, but we couldn't make it affordable for most homes.
For sure! I drive right through the mountains in the ADK often. Unless you're in a well populated village, there is no cell service at all. Break down on the road and you're walking 10+ miles.
That also leads to slippery-slope politics with the Adirondack Park Agency as to why cellsite adoption has been so slow and frustrating.
Speaking as someone who used to live in Long Lake and had to drive to the edge of the hamlet down at a bridge just to get a single bar to check back in on family, I certainly wish the APA would loosen up a bit more on just letting towers go up.
Thanks to unfavorable geography I get no cell phone service where I live. My power lines go through a beaver swamp so conditions are highly unfavorable for FTTH. (The power company does go through every few years in an off-road vehicle to limb up the trees and if we had a realistic commitment to universal service we'd just run fiber everywhere the power lines go and it would be taken care of eventually)
I have 2 ADSL2 lines that connect to a fiber node a bit more than a mile away, each gets about 20 Mbps of bandwidth. Spectrum gives a lot more bandwidth for less money just 2.5 miles away, but I really get 100% of what I am paying for almost all of the time so I can't complain too much.
Look at how conservatives would howl if there wasn't a universal mandate for electric power everywhere. I mean, if we weren't entitled to energy from coal we'd have to put up solar panels.
Had we really been thinking straight about the "information superhighway" in 1994 we would have just decided it was a function of the electric company and strung an optic fiber along with the power lines and it would have been "least cost".
People writing about the future in the 1970s expected we would go straight to fiber without detours into CATV, DSL, spotty, expensive and low-performing wireless systems, etc.
Neither Verizon or T-Mobile's home internet products have data caps. In fact, of the four ISPs my home has access to, Comcast Xfinity is the only one that has a cap.
How consistent is the latency? My friend has used T-Mobile 5g for a few months and while it is much better than the other options it still has some issues (massive random ping spikes often 1000+ms). He is hoping the issue gets better in the future.
It goes to hell if you max out upstream. At cell edge it's easy to do.
Modem type in use greatly affects how it'll perform real-world. N71 (600mhz) standalone 5G without carrier aggregation tends to be shit as just like Band 12 (700mhz) LTE. It's their low frequency coverage band for 5G and congests since the towers are spaced far apart and it uses less spectrum. Newer phones can aggregate multiple N71 and N41 carriers and get very low latency high performance connections, but it's still location dependent.
It appears to be entirely at T-Mobile's discretion; they reserve the typical mobile carrier rights to shape any traffic they find unfavorable. I expect you will see the usual mobile industry reductions in throughput once you pass an arbitrary line each month. While I love the theoretical performance, until that performance is documented and guaranteed I am out.
"You can use your Home Internet for all the homework, streaming, and video conferencing you need! But to ensure that our network is available for all customers, there are some restrictions on activities that can damage or disproportionately congest the network. For example, Home Internet is not intended for unattended use, automatic data feeds, automated machine-to-machine connections, or uses that automatically consume unreasonable amounts of available network capacity."
I think its fair to assume my definition and T-Mobile's definition of "unreasonable" will not be in alignment.
Anecdotally, I have been using multiple TBs/month of transfer with zero issue or throttling. I'm in HD Zoom calls 4-5hrs a day and do heavy gaming, torrenting, and streaming.
So much of what you receive is going to fall down to classic contention to backhaul ratios too (how many people connected to your tower, sharing a fixed amount of bandwidth etc). Just the presence of the "5G" signal tells you not much about how it will actually perform in your own neighborhood vs somewhere else.
I would still take a substantially slower but reliable copper connection over this service, but thats a matter of personal preference. FWIW, you don't need to search too hard to find evidence of throttling on T-Mobile's cellular network. I still think the concept is great and the idea has merit though, it's just early days in the "unlimited" wireless home internet market.
All high-bandwidth wireless technologies are limited to line of site or near line of site. In the U.S. from the Southern Tier of NY down to Georgia and Alabama many people live in a landscape which is basically a plateau which is dissected by valleys.
There are no "commanding heights" in these areas, rather a tower on a ridge between two valleys may be able to see into those valleys and maybe the next two over.
Sometimes the top of the ridge is roadless so putting in wireless infrastructure is a matter of building a new road, a new power line, running fiber optic along the power line, etc. It boggles the mind to think that this can possibly cost less than running fiber along existing roads and power lines (either for FTTH or FTTN) but that's what the anti-fiber lobby wants people to believe.
If you're in some place where 5G is available it is probably a relatively high-density place that already has cable, in which case 5G internet service is usually trying to undercut cable on price. If you're in a place where WISP is a substantial option you're either in an unusually favorable urban location (the webpass model) which is vulnerable to millimeter-wave 5G or in a a rural location where the WISP is already a desperate measure and 5G won't be profitable.
A partial counter is that AT&T and T-Mobile have added a lot of low band spectrum in recent years that have some potential in Appalachian environments and served it with either 4G or 5G radios but the spectrum does not add up to a lot of bandwidth, particularly if radio conditions are unfavorable.
Why would you need a road? The equipment isn't that big, a few men, maybe a donkey, could put a base station up in a few days. The cable can be strung along trees to it.
Maybe for a base station that's just sitting on the ground. But to actually be useful, wouldn't it need an antenna mast that's taller than the surrounding trees?
There is a base station on top of my hill (which has a road) but cell phone coverage seems poor to nonexistent.
The base station involves large gravel pad, a solid (as opposed to space frame) tower that must have required a crane to erect, a large multi-sector antenna, an equipment building about the size of a singlewide mobile home, a diesel generator and a chain link fence.
Mao's Red Army might have moved that much with a mule train in the 1930s but I don't think anybody is doing that without a road today.
And by my house there's two different companies sharing a tower with all non-pole-mounted equipment in an 8'x8' shed plus generator. You could have the generator and batteries by the nearest road instead of at the antenna site.
But that would require a non-oligopoly with cost pressure and lack of regulatory capture.
Would a Sigorsky be feasible? Not an idle question, I type this from a remote Aleutian island in sight of a 4000+ foot peak and no infrastructure to get a tower there.
It’s worth remembering though that new generations of mobile networks typically have great performance in the beginning, when few people are using them. But performance quickly deteriorates when more and more subscribers are onboarded and the competition for spectrum access begins.
I was looking at a Wisp for a rural weekend home and it was 15x15Mbps for $50/month for unlimited plus $200 for the equipment on install. When I looked at t-mobile 5G home internet it was $50 for unlimited and free equipment. Although the speeds fluctuate they are still faster than the wisp.
I think It's going to be mid range 5G that kills wisp, they might be able to serve edge cases but the mobile carriers see a market for home internet. They have the technology and the spectrum to make it happen quickly and everyone already has a mobile phone.
Not surprising. I did my thesis on community run broadband networks and the handful of WISPs I studied all had really pitiful offerings (comparatively speaking). Kind of sad, but the writing was on the wall with 4G being generally available in the US.
WISPs are such a cool idea, but I'm glad to see fiber and 5G being rolled out. Not a huge fan of mobile providers gaining bigger market shares but that's the nature of the beast.
Our local WISP nailed the last few mile problem, but zero providers will sell them more bandwidth for their towers, so we get 0.5s latency, 0.1M down, and single digit percent packet loss a hop or so past the tower. (The wireless link is reliably 30/30, near speed of light latency)
T-Mobile keeps falsely advertising the availability of uncapped 5G at my address (when I try to buy it after they text me saying unlimited is available at my home address, the price more than doubles for a low data cap).
>McGrath said in the territory he serves in Texas he’s received a number of requests from parents looking for a completely different connection for their children.
Or just get a modem with parental control? Considering that he used that as an example of how WISPs can stay relevant even with more fiber... I'm not convinced.
Also, I always imagined WISPs to be very local and only for very remote communities. But they mentionned some providers having up to hundreds of thousands of clients in a single state. How bad does the service get at that point?
Even wired docsis 3.1 can be pretty horrible in terms of reliability, latency, congestion, etc so I'd guess the congestion issue is even worse for wireless internet?
58 comments
[ 4.7 ms ] story [ 268 ms ] threadWe have a solution!
Solar powered Internets!
Sun is down - no Tiktoks around!
/s, but thanks for a laugh.
https://en.m.wikipedia.org/wiki/Power-over-fiber
I have the T-Mobile 5G gateway as my sole home internet connection. I get a rock solid 300/100 with 30ms latency for $50/month. That approaches DOCSIS 3.1 performance, let alone WISP performance. Not sure how they will be able to compete.
Around there this is why they are decommissioning all copper lines. There isn’t a market left for it.
So now you either have 4G/5G or fiber. Sucks if you don’t have fiber.
I run a weather station on T-Mobile. Yes it's ipv6-only inbound but a Cloudflare free plan is enough to provide legacy (ipv4) access to the station. While I agree that running a media server wouldn't be possible from this location, I disagree that it wouldn't work closer to a tower. Upload speeds there are over double what I can buy on Comcast business's most expensive plan.
That said, hosting from home and syndicating to a VPS is a great idea.
5G is just the nail in the coffin.
I never thought of WISPs existing elsewhere as it doesn't make nearly as much sense if you have access to "real" broadband. I suppose this niche was wider at one point?
Speaking as someone who used to live in Long Lake and had to drive to the edge of the hamlet down at a bridge just to get a single bar to check back in on family, I certainly wish the APA would loosen up a bit more on just letting towers go up.
That isn't entirely true, G.fast is a pretty decent alternative here, unlimited 4G/5G is too unreliable for a lot of people.
I have 2 ADSL2 lines that connect to a fiber node a bit more than a mile away, each gets about 20 Mbps of bandwidth. Spectrum gives a lot more bandwidth for less money just 2.5 miles away, but I really get 100% of what I am paying for almost all of the time so I can't complain too much.
Had we really been thinking straight about the "information superhighway" in 1994 we would have just decided it was a function of the electric company and strung an optic fiber along with the power lines and it would have been "least cost".
People writing about the future in the 1970s expected we would go straight to fiber without detours into CATV, DSL, spotty, expensive and low-performing wireless systems, etc.
WISPs should be quite safe in the USA, then; data caps are still everywhere. Actually, I had a data cap even on wired fiber, just higher.
Modem type in use greatly affects how it'll perform real-world. N71 (600mhz) standalone 5G without carrier aggregation tends to be shit as just like Band 12 (700mhz) LTE. It's their low frequency coverage band for 5G and congests since the towers are spaced far apart and it uses less spectrum. Newer phones can aggregate multiple N71 and N41 carriers and get very low latency high performance connections, but it's still location dependent.
2. It's a wireless tech, so you would be affected if the guy 30m from you decided to download all Internets.
What, if any, are the caps on that service?
A quick look at the T-Mobile site doesn't say one way or the other.
"You can use your Home Internet for all the homework, streaming, and video conferencing you need! But to ensure that our network is available for all customers, there are some restrictions on activities that can damage or disproportionately congest the network. For example, Home Internet is not intended for unattended use, automatic data feeds, automated machine-to-machine connections, or uses that automatically consume unreasonable amounts of available network capacity."
I think its fair to assume my definition and T-Mobile's definition of "unreasonable" will not be in alignment.
https://www.t-mobile.com/home-internet/faq
Anecdotally, I have been using multiple TBs/month of transfer with zero issue or throttling. I'm in HD Zoom calls 4-5hrs a day and do heavy gaming, torrenting, and streaming.
I would still take a substantially slower but reliable copper connection over this service, but thats a matter of personal preference. FWIW, you don't need to search too hard to find evidence of throttling on T-Mobile's cellular network. I still think the concept is great and the idea has merit though, it's just early days in the "unlimited" wireless home internet market.
All high-bandwidth wireless technologies are limited to line of site or near line of site. In the U.S. from the Southern Tier of NY down to Georgia and Alabama many people live in a landscape which is basically a plateau which is dissected by valleys.
There are no "commanding heights" in these areas, rather a tower on a ridge between two valleys may be able to see into those valleys and maybe the next two over.
Sometimes the top of the ridge is roadless so putting in wireless infrastructure is a matter of building a new road, a new power line, running fiber optic along the power line, etc. It boggles the mind to think that this can possibly cost less than running fiber along existing roads and power lines (either for FTTH or FTTN) but that's what the anti-fiber lobby wants people to believe.
If you're in some place where 5G is available it is probably a relatively high-density place that already has cable, in which case 5G internet service is usually trying to undercut cable on price. If you're in a place where WISP is a substantial option you're either in an unusually favorable urban location (the webpass model) which is vulnerable to millimeter-wave 5G or in a a rural location where the WISP is already a desperate measure and 5G won't be profitable.
A partial counter is that AT&T and T-Mobile have added a lot of low band spectrum in recent years that have some potential in Appalachian environments and served it with either 4G or 5G radios but the spectrum does not add up to a lot of bandwidth, particularly if radio conditions are unfavorable.
The base station involves large gravel pad, a solid (as opposed to space frame) tower that must have required a crane to erect, a large multi-sector antenna, an equipment building about the size of a singlewide mobile home, a diesel generator and a chain link fence.
Mao's Red Army might have moved that much with a mule train in the 1930s but I don't think anybody is doing that without a road today.
But that would require a non-oligopoly with cost pressure and lack of regulatory capture.
http://www.hydrogen18.com/blog/trying-tmobile-5g-internet.ht...
I think It's going to be mid range 5G that kills wisp, they might be able to serve edge cases but the mobile carriers see a market for home internet. They have the technology and the spectrum to make it happen quickly and everyone already has a mobile phone.
WISPs are such a cool idea, but I'm glad to see fiber and 5G being rolled out. Not a huge fan of mobile providers gaining bigger market shares but that's the nature of the beast.
T-Mobile keeps falsely advertising the availability of uncapped 5G at my address (when I try to buy it after they text me saying unlimited is available at my home address, the price more than doubles for a low data cap).
Starlink RV plan for me for now. sigh
It's OK, but ridiculously over congested.
Or just get a modem with parental control? Considering that he used that as an example of how WISPs can stay relevant even with more fiber... I'm not convinced.
Also, I always imagined WISPs to be very local and only for very remote communities. But they mentionned some providers having up to hundreds of thousands of clients in a single state. How bad does the service get at that point? Even wired docsis 3.1 can be pretty horrible in terms of reliability, latency, congestion, etc so I'd guess the congestion issue is even worse for wireless internet?