FCC releases report on America's broadband providers | Hacker News Reader

FCC releases report on America's broadband providers

fcc.gov

54 pointsecaron22 comments

Nice report with numbers for my provider matching what I've seen. However, there is no mention of bandwidth caps, internet downtime, or availability of competing services. These three factors should be reported on as well.

Agreed, very narrow focus and although the results are positive, the publication does little to spur on development in this industry. American ISPs, we need to talk.

Agreed. Downtime in NYC from TWC is painful. Youtube actually has a tool that shows much of the information available in this report for your own connection: http://www.youtube.com/my_speed

Strangely enough, in the FCC's wireless report, they could not come to a conclusion about the state of competition in the wireless industry. I would imagine that coming to a conclusion about wireline broadband would be equally (if not more) difficult for them.

> Downtime in NYC from TWC is painful

Really? I run nagios at home, and right now it says year-to-date I've had 28m 40s of downtime, for 99.988% uptime of my cable internet service (for the host monitoring my first-hop router). This is in Chelsea, Manhattan, with their "wideband" level of service.

I'm actually doing some research on this topic right now in grad school. I think there is definitely a lot more that can be done in this area; we're looking at things like time-of-day congestion, traffic management policy, actual application performance, and reliability. I think, as you said, there is a lot more to this than just bandwidth measurements. For example, last November a bunch of DNS servers went down in Comcast's networks that essentially meant no Internet for the majority of people that don't know how to configure public DNS.

We're working to eventually get a much finer grain survey of performance (e.g. a neighborhood), by piggy backing on network-intensive applications (currently BitTorrent). In addition to throughput, we're looking at things such as packet loss, latency (the last-mile in particular), time-of-day effects, as well as using traceroutes for underlying forward-path analysis.

My point is, if anyone has any ideas for other applications (something like Netflix would be great) to leverage, metrics to include, or advice, I'd be happy to hear them. :)

To frame this another way: if we only use our connections for short periods of time, we can almost get 100% of what we're paying for. Woo!

Cablevision seems to be the largest offender, consistently delivering ~69% of advertised speed. Qwest is another I noticed from skimming the data that consistently underperforms by a significant margin.

If anything, this reads like an advertisement for FiOS. Verizon's FiOS dominates every (positive) metric. I've heard that they've stopped deploying FiOS to new districts. A shame, if true.

Guess I can always hope Google's fiber project goes well and someday sees deployment in the Chicago-land area.

Page 25. It reflected perfectly my experiences with Cablevision: I can stream movies anytime but prime-time.

And it's the only provider I can get other than DSL. Ugh.

I was hoping to see some measurement of dropped connections - something that I used to see sometimes when I was still using DSL.

I skimmed the report, but one point caught my attention:

VoIP services, which can be used with a data rate as low as 100 kilobits per second (kbps)

Isn't this quoted number a little high? If I remember correctly, Ventrilo on a good codec was only 4kb/person/second.

That's true. Modem was at 56kps max mostly on regular phone line. Voice requires much less than 56kps. Voice on old line sounds crisp and clear.

Are they planning for mulit-VoIP sessions or video VoIP?

A POTS line is analog so even though you can run a 56kbps modem, the actual data is transmitted via an analog signal. G.711 is the raw digital voice benchmark and encodes at a constant 64kbps PCM a-law or u-law ecoding (8 kHz sampling frequency x 8 bits per sample).

This is the reason why digital carrier lines are always multiples of 64. A DS-0 is one 64kbps channel, a DS-1 is 1.54 Mbps (24 * 64), an E1 is 2.048 Mbps (32 * 64), etc.

For low latency, the IP/UDP/RTP headers are much bigger than the audio in many cases. So a "64 kbps" codec is actually ~96 kbps (IIRC) and the overhead gets worse at lower bitrates.

G.711 (no compression) is 64kb/sec. G.729* is 8kb/sec.

The real problem with VoIP is you need low latency. I believe 80ms is where you can start getting into the area where people notice the delay.

* EDIT: I forgot to mention that G.729 is encumbered by a number of patents.

I'm actually a part of this survey. I have a SamKnows device (basically a Netgear router with custom firmware) that polls Charter's 'net service and I get monthly details and can even login to the SamKnows site to see how Charter is doing. Pretty slick.

I wonder if there is a legal requirement for an ISP to provide a certain % of advertised performance. Also wonder if ISP circumvents by using "Up to" in advertising. up to 3 MBPS usually comes out to 56kps 99% of the time in my case.

The report matches my experience with Verizon FIOS. By far the best internet service in America. I was coming from TWC and had positive experiences there, but FIOS is even better. Low latency, high bandwidth.

Same with Cox and me. Pretty fast speeds (for the South and cable, anyway), great ping, and no caps.

Hate to ask, but is there a way to make a note of a thread for later, aside from making a comment in it?

Yes, that is what I'm doing here. On my phone right now.

Upvoting it will put it in the "saved stories" section in your profile, but that's about it.

Pimping my tool:

http://github.com/bsandrow/hn-profile

Written in response to an HN thread asking how to download the list of saved articles.

Ahhh, thanks to both of you!

Reply on news.ycombinator.com