Hacker News Reader

Comment by Cshelton | Hacker News Reader

The future of Telco is not in the ground, but through the air. This is the entire reason Google Fiber acquired WebPass.

It is relatively simple to transmit extremely high connection speeds through radio waves. I have a receiver on the top of my house that is a point to point signal, pumping 1 gig +, and soon to be much higher, speeds.

Google's strategy to expand GF will not longer include a massive ground infrastructure move, but use point to point to get to the neighborhoods. An entire zip code can be offered GF services at a fraction of the cost and a few months of labor. Google, along with many other broadband startups will hit the big telco's very hard. I'm not even sure if they see it coming yet. I'd imagine they would, but after dealing with AT&T for the last two months on something, I'm not very confident in there being any single person in the entire organization who has a clue about what they are doing.

View on HN

You aren't getting gigabit class service to "an entire zip code" using only wireless. There are hard physical limits to how radio waves propagate and how many bits you can squeeze into a given amount of spectrum [0]. Radio links can provide high bandwidth or long-distance and reasonable obstruction penetration but the former doesn't really come with the latter. Range and obstacle penetration come in the sub-GHz class radios, but there isn't much bandwidth available.

[0] https://en.wikipedia.org/wiki/Shannon%E2%80%93Hartley_theore...

Microwave seems promising.

Microwave, as a radio term, isn't very useful. It covers 300 MHz to 30 GHz [0], which is nearly the entire usable radio spectrum. In industry terminology, microwave generally refers to >GHz radios, which have real problems in terms of obstacle penetration and range. In the >GHz range, you more or less require line of sight for transmissions over a kilometer or so, which often isnt practical. Free space path loss [1] is also an issue on long links, as signal loss is proportional to the square of frequency.

A good practical example is to compare 2.4GHz WiFi to 5GHz WiFi. At the same power level (both are restricted to 1W in the US for unlicensed operation), 2.4GHz will cover approximately four times the distance ~(5/2.4)^2. However, much more spectrum is available in the 5GHz range, and signals cant travel as far which leads to less interference given an equal number of transmitters. Those factors give higher throughput, but free space path loss constrains coverage on 5 vs 2.4 GHz WiFi.

Another example would be cellular: low frequency towers are used to expand coverage and improve building penetration, high density high frequency towers are used to improve speeds.

Feel free to email me at my profile address -- I work in the sub-GHz narrowband radio industry.

[0] https://en.wikipedia.org/wiki/Microwave [1] https://en.wikipedia.org/wiki/Free-space_path_loss

I'm talking about site to site.

I'll clarify, sorry. To provide 1 gig+ service to an entire zip code, it would be one transmitter and one receiver, point to point. The receiver would then have a distribution box and run cables and/or other smaller wireless to surrounding neighborhood's/houses. The cost of doing it this way is significantly less expensive and quicker to bring service to many more houses. This will be Google Fiber's model as they expand into many more cities. The entire reason they bought WebPass.

We're using microwave-based internet in SJ, with Ridge Wireless. Our experience has been that speeds are slow and whenever a cloud passes between our transceiver and the remote transceiver, or if it's a rainy day, etc, speeds suffer really badly.

I don't know much about WebPass, but I'm not sure that the future is in the air. I'm not convinced that companies have yet built up the infrastructure to e.g. monitor customer link quality, and deal with issues like weather, antenna orientation, etc proactively. The beauty of a cable in the ground is that it pretty much stays the same over time, at least relative to antennas.

I would really love to hear more about how WebPass manages these things.

I'm a current WebPass subscriber and I'm dreading the move to another ISP when I change apartments. May I inquire what it takes to set up and maintain the transmitter?

I believe the transmitters are pretty expensive. In the $10,000 range, depending on the frequency and power required.

So there's cost if you want to own one. Plus you need to convince someone (Webpass I guess?) to point one of theirs at you to complete the link.

It's probably easiest to convince the owner of your new apartment to pay Webpass's setup fee, where they own the equipment and wire the building. Although having tried that a couple of times, it's a pain in the ass.

Perhaps because of this: https://backchannel.com/the-new-payola-deals-landlords-cut-w...

The receivers themselves for a consumer you can get for < ~$1,000 for one that will be able to handle a couple of hundred Mbps easy. Getting someone else to point directly at you would cost a lot. Or doing both ends yourself. The business model for ie. WebPass is not to point to every customer, but to point to one point in say a neighborhood and then run cables from that to every house in the neighborhood. Which is a fraction of the cost of running cables all over cities.

At Least, that will be the business model under Google Fiber.

The biggest requirement is to have line of sight to the headend or some other already connected building.

This is simply ignorant of physics.

Please explain?

I currently have this set up and operating. Infact, we have one transmitter transmitting to 3 separate, close proximity buildings. To call it ignorant when this is what we are already doing is odd.

Reply on news.ycombinator.com