4 ms·

When the "fast wireless" under discussion more or less requires deploying FTTC as the backhaul, consumers definitely stand to benefit from transforming that int

When the "fast wireless" under discussion more or less requires deploying FTTC as the backhaul, consumers definitely stand to benefit from transforming that into a regulated FTTH network instead of being further subjugated by cellular carriers.

tptacek 7y ago
Isn't the cost difference between running fiber to base stations and running fiber to individual homes enormous? Just based on the economics: a fiber-lit base station serves anyone nearby with a mobile device, while the infrastructure to support FTTH is a risk (and, apparently, an unprofitable risk) based on how many homes sign up. I don't think base station backhaul and FTTH are comparable investments.
- wtallis 7y ago
  If we're talking typical suburban housing density, deploying 5G mmWave will require digging trenches or hanging wires on poles through almost as many properties as deploying FTTH. A 5G basestation is going to be a lot more expensive than a GPON ONT, probably enough that the per-customer costs aren't a net win for 5G. (When the fiber is already at the mailbox, taking it down the driveway too just isn't that much more investment.) And the 5G basestation faces the same problem as FTTH that not every person in the neighborhood will be a guaranteed subscriber to any particular provider; even if Verizon/ATT/etc. already has you under contract for your smartphone, they have to work on convincing you to add all your wired devices to the plan. Barring detailed cost estimates to the contrary, I don't see how mmWave will be viable in suburban America. Sub-6GHz 5G will be, but that's not going to be providing dramatic speed upgrades or price/Mbps improvement the way FTTH has. In a denser urban residential setting, mmWave certainly won't have enough penetration to provide solid service to high-rise apartments, and it's even less likely that sub-6GHz 5G will have enough aggregate capacity to kill off wired ISPs. mmWave simply isn't viable for residential/fixed usage, and sub-6 is only an incremental improvement that brings cellular performance up to WiFi speeds, but not beyond.
  - rayiner 7y ago
    > If we're talking typical suburban housing density, deploying 5G mmWave will require digging trenches or hanging wires on poles through almost as many properties as deploying FTTH. This is far off base. Conservatively, a mmwave base station has a range of about 1,500 feet, covering about 7 million square feet, or 160 acres. Median subdivision density is about 2 houses per acre: https://www.nahbclassic.org/generic.aspx?sectionID=734&genericContentID=253886 https://www.nahbclassic.org/generic.aspx?sectionID=734&gener.... So one base station can cover 300 houses. For comparison, an N+0 cable node (a node where the fiber is close enough to the houses to require no amplifiers on the coax side) is 50-130 houses. Cable companies have been running N+0 nodes all over the place, but have been hesitant to build the fiber all the way to the premises. That should tell you a lot about how expensive that next increment is.
    - wtallis 7y ago
      > Conservatively, a mmwave base station has a range of about 1,500 feet, At what speed? If that's just the range at which it stops being an upgrade over current LTE speeds, that's really unimpressive. The figure of interest here is the range at which mmWave can reliably sustain WiFi-like performance, including enough aggregate capacity for gaming to still have low latency when all your neighbors and one or two of your family members are streaming video. > That should tell you a lot about how expensive that next increment is. Not really. Cable companies are upgrading their networks by replacing some of the coax in the ground with fiber, as necessary to get enough backhaul bandwidth to each node. The incentives there are wildly different than for a new ISP that's either deploying FTTH or short-range cellular, without having access to pre-existing last mile wiring that has no trouble carrying gigabit to each customer. For wired infrastructure, digging the trenches is a lot more expensive than the materials cost for whatever you're putting in the trenches.
      - rayiner 7y ago
        Testing of the Houston Verizon deployment is showing near a gigabit at 2,000 feet: http://wirelessone.news/8-high-frequency-mmwave/981-2-000-feet-verizon-5g-field-test-nlos-often-close-to-a-gig http://wirelessone.news/8-high-frequency-mmwave/981-2-000-fe... As to your second point, the whole point of wireless is you avoid the need to dig trenches. Getting the last 500 feet through a subdivision is a huge pain in the ass. I had Comcast install fiber to my house a couple of years ago. It took the contractors half a day to run fiber 1,700 feet on the pole down the main road, another half a day to run it 300 feet to the pole outside my house, and another half day to run it 30 feet under my driveway. With 5G they could’ve stopped after the first step and served the whole subdivision (200 houses).

tptacek 7y ago

Isn't the cost difference between running fiber to base stations and running fiber to individual homes enormous? Just based on the economics: a fiber-lit base station serves anyone nearby with a mobile device, while the infrastructure to support FTTH is a risk (and, apparently, an unprofitable risk) based on how many homes sign up. I don't think base station backhaul and FTTH are comparable investments.

wtallis 7y ago
If we're talking typical suburban housing density, deploying 5G mmWave will require digging trenches or hanging wires on poles through almost as many properties as deploying FTTH. A 5G basestation is going to be a lot more expensive than a GPON ONT, probably enough that the per-customer costs aren't a net win for 5G. (When the fiber is already at the mailbox, taking it down the driveway too just isn't that much more investment.) And the 5G basestation faces the same problem as FTTH that not every person in the neighborhood will be a guaranteed subscriber to any particular provider; even if Verizon/ATT/etc. already has you under contract for your smartphone, they have to work on convincing you to add all your wired devices to the plan. Barring detailed cost estimates to the contrary, I don't see how mmWave will be viable in suburban America. Sub-6GHz 5G will be, but that's not going to be providing dramatic speed upgrades or price/Mbps improvement the way FTTH has. In a denser urban residential setting, mmWave certainly won't have enough penetration to provide solid service to high-rise apartments, and it's even less likely that sub-6GHz 5G will have enough aggregate capacity to kill off wired ISPs. mmWave simply isn't viable for residential/fixed usage, and sub-6 is only an incremental improvement that brings cellular performance up to WiFi speeds, but not beyond.
- rayiner 7y ago
  > If we're talking typical suburban housing density, deploying 5G mmWave will require digging trenches or hanging wires on poles through almost as many properties as deploying FTTH. This is far off base. Conservatively, a mmwave base station has a range of about 1,500 feet, covering about 7 million square feet, or 160 acres. Median subdivision density is about 2 houses per acre: https://www.nahbclassic.org/generic.aspx?sectionID=734&genericContentID=253886 https://www.nahbclassic.org/generic.aspx?sectionID=734&gener.... So one base station can cover 300 houses. For comparison, an N+0 cable node (a node where the fiber is close enough to the houses to require no amplifiers on the coax side) is 50-130 houses. Cable companies have been running N+0 nodes all over the place, but have been hesitant to build the fiber all the way to the premises. That should tell you a lot about how expensive that next increment is.
  - wtallis 7y ago
    > Conservatively, a mmwave base station has a range of about 1,500 feet, At what speed? If that's just the range at which it stops being an upgrade over current LTE speeds, that's really unimpressive. The figure of interest here is the range at which mmWave can reliably sustain WiFi-like performance, including enough aggregate capacity for gaming to still have low latency when all your neighbors and one or two of your family members are streaming video. > That should tell you a lot about how expensive that next increment is. Not really. Cable companies are upgrading their networks by replacing some of the coax in the ground with fiber, as necessary to get enough backhaul bandwidth to each node. The incentives there are wildly different than for a new ISP that's either deploying FTTH or short-range cellular, without having access to pre-existing last mile wiring that has no trouble carrying gigabit to each customer. For wired infrastructure, digging the trenches is a lot more expensive than the materials cost for whatever you're putting in the trenches.
    - rayiner 7y ago
      Testing of the Houston Verizon deployment is showing near a gigabit at 2,000 feet: http://wirelessone.news/8-high-frequency-mmwave/981-2-000-feet-verizon-5g-field-test-nlos-often-close-to-a-gig http://wirelessone.news/8-high-frequency-mmwave/981-2-000-fe... As to your second point, the whole point of wireless is you avoid the need to dig trenches. Getting the last 500 feet through a subdivision is a huge pain in the ass. I had Comcast install fiber to my house a couple of years ago. It took the contractors half a day to run fiber 1,700 feet on the pole down the main road, another half a day to run it 300 feet to the pole outside my house, and another half day to run it 30 feet under my driveway. With 5G they could’ve stopped after the first step and served the whole subdivision (200 houses).