Broadband isn't Broadband Unless its 2Mbps?

quanticle writes "According to House Democrats, broadband isn't broadband unless its at least 2Mbps. The view of the House Subcommittee on Telecommunications is that the FCC's data collection standards are hopelessly outdated, and is proposing a number of updates to their criteria. For one, they want 'broadband' reclassified to at least 2mbs, up from 200kbps. Another requirement will change the FCC's outlook on broadband availability. Just because one household in a zip code has broadband access, that will not longer mean everyone in the zip code does. 'The plan went over well with the consumer advocates who appeared before the subcommittee. Larry Cohen, president of the Communication Workers of America, said that the US is "stuck with a twentieth century Internet" and that he would support increasing the "broadband" definition to 2Mbps. Ben Scott of Free Press echoed that sentiment, suggesting that the definition needs to be an evolving standard that increases over time, which is in contrast to the current FCC definition; it has not changed in nine years. "We have always been limited by the FCC's inadequate and flawed data," he said.'"

Re:Korea has 10MBPs to the home...

[zappepcs]

10Mbytes? Why stop there? Fiber will give you 30 easily. The infrastructure upgrade to handle all those 30Mbyte end user connections, but that can be done over years. It won't be long before wireless will be competing successfully with DSL and making dialup seem a bad value.

What again was broadband?

[dascandy]

Broadband, as opposed to baseband, is technically defined as anything not at the base frequency of 0Hz. Baseband is at the base frequency and up, broadband is at a higher frequency and up.

FCC can't even seem to get a technicality right.

Re:What about uplink speed?

[grub]

Far too many people are stuck on lines that have 128Kbps up and far too easily saturate the uplink and bog the whole connection down.

That's why it's handy to have a decent gateway which can prioritize TCP ACKs. If they get lost in the muddle your download speeds get hurt. It's covered here [openbsd.org]. (I link to the OpenBSD pages as that's what I use)

768k

[WatchTheTramCarPleas]

768K seems to be a nice low speed broadband. Large downloads are still doable, and youtube videos just take a few more seconds to buffer than on a faster connection. Podcasts are downloaded automatically in the background, so there is little reason for those to have to be super fast. This is just to serve as an example of working broadband internet under 2mb.

Re:What about uplink speed?

[aegzorz]

They do the same thing here in Sweden, they say you'll get 24Mbit broadband (DSL) but the uplink is only 1Mbit. Most people only use the Internet for webbrowsing but more and more use it for VoIP, 1Mbit up is awfully slow when you use services like that.

I currently have a 100/100Mbit Internet connection, but they're offering up to 1Gbit in other parts of my city. They won't really get 1Gbit but certainly somewhere around 400Mbit.

For it to be called broadband I think the bandwidth should have to be symmetric, or at least 2:1.

Look out for the fine print

[azrider]

It's nothing to do with that; it's just regulating what the cable/phone companies can claim actually IS broadband; as it stands they screw a lot of people who don't know any better by selling them "broadband" which is no such thing by modern standards. I think it's definitely a good step in the "truth in advertising" department...I'm tired of sneering at the commercials where the broadband companies are comparing their download speeds to 28.8 modems and other such crap.

Just look at some of the offerings. One that I am familiar with advertises 3 different wireless services (768k MIR for $59, 1M MIR for $99, 3M MIR for $139). MIR stands for Maximum Information Rate as in "Up to 3Mb/sec". However, each of the services also specifies a CIR (Committed Information Rate) of 512k. This means that, until your rate drops below 512k/sec, you cannot complain that they are not adhering to their part of the contract.

Remember that and always ask for both the CIR and MIR when talking to a sales person. If they will not specify a CIR (or don't know what it is), RUN, don't walk for the nearest exit.

Re:LLU's dead; the FCC killed it.

[Anonymous Coward]

The phone company still has to sell local loops. What was taken away was having to share a loop that they were providing service over. So you can get things like Speakeasy's Oneline service, it just costs about $10/month more than if it was provided on the same loop as you local phone service.
The other stupid change that was made at some point was not allowing CLECs to share the fiber from a CO to a remote terminal. They can get access to the copper loops going from the terminal to people's homes, but they have to pay a lot of money to run their own fiber to make the copper of any value to them.

Baseband IQ

[dunc78]

I have never heard these definitions to which you are referring. I have heard definitions similar to your baseband defintion, with the difference being that baseband signals are complex signals CENTERED at 0 Hz, not signals going from 0 to some other frequency. The terminology I have heard to refer to a signal going from 0 to F1 would be, a 200% bandwidth signal at a center frequency of F1/2. I have never heard anything remotely similar to your broadband definition. Broadband is a relative "bandwidth" term and has nothing to do with the center frequency. I would be curious to hear if other people have heard your definitions, if not I would say it is you that is techically wrong.

Re:Baseband IQ

[Anonymous Coward]

It depends on how you define the frequency spectrum and which corresponding set of transforms you use.

If you define frequency spectrum from (-inf,inf), meaning you can have negative frequency components then:
Baseband is centered at 0 from [-freq, freq].
Broadband is centered at 0 from the union of [-freqHIGH, -freqLOW] and [freqLOW, freqHIGH].
And you need to use double-sided transforms.

If you define frequency spectrum from [0,inf), meaning you have only absolute frequency components then:
Baseband is centered at freq/2 from [0, freq].
Broadband is centered at (freqLOW + freqHIGH)/2 from [freqLOW, freqHIGH].
And you need to use single-sided transforms

It's all the same in the end as long as you are consistent using the corresponding mathematical toolset for
single or double-sided transforms.

Re:Look out for the fine print

[Jah-Wren Ryel]

Remember that and always ask for both the CIR and MIR when talking to a sales person. If they will not specify a CIR (or don't know what it is), RUN, don't walk for the nearest exit.

Last I checked, all of the major wired broadband providers in the USA had a CIR of 0bps for their consumer-grade services. So, either you run away, get no connection at all and never exceed that CIR of 0bps or you pick the least worst and usually get significantly more than the CIR.