Broadband or not

There is discussion in some newsgroups (and other places) about what is and isn't broadband.

It seems to me that the meaning, in EE terms, relates to the bandwidth relative to the carrier frequency, where wider relative bandwidth requires more complicated filters.

In ethernet terms, though, it seems to be the antonym of baseband, which seems to imply no narrowband ethernets.

In addition, it seems to me that the current cable modem systems should be a form of broadband ethernet, though they are not usually described that way. (And even more, at the high carrier frequencies they may not even qualify as broad.)

-- glen

Reply to
glen herrmannsfeldt
Loading thread data ...

Here in the US, connected to CATV systems, the LAN side is 100Mb/S while the WAN side is only 6-7Mb/S (uploads are only a bit over

1Mb/S). We're supposed to think this is Broadband - only because it's fatter than DSL or dialup.

I'm working for a WiMAX provider, and customer connections are usually capped similarly, but at least they're mobile and competitive. My uncapped device can get a lot fatter speeds, but that's just for testing purposes. The point is, the tech works well.

If all the providers think IPTV is the "next big thing" they're going to have to rethink their idea of what broadband is...

On Wed, 21 Jul 2010 20:40:38 +0000 (UTC), glen herrmannsfeldt wrote (with clarity & insight):

Reply to
Wayne R.

Or it could more simply imply, "broader passband than 4 KHz POTS bandwidth," used previously for Internet connections to homes.

I'm not sure why the "relative to the carrier frequency" should necessarily be included in the meaning. If you look at the various schemes out there, carrier frequencies are all over the map.

The problem with cable modems is that they are connected to a non- Ethernet cable plant. Cable modems are connected to a coax plant that is passively split among several homes, just like a TV antenna downlead might be, when connected to multiple sets.

The downstream Internet traffic uses one RF channel, or more possible with the newest schemes, and the upstream use a different RF channel. So, this means that in order to go from one house to another within your neighborhood, the traffic needs to go to the coax headend first, then back down. Even if the two houses are next to each other.

So as you can see, this is neither like a 10BASE2 Ethernet (where hosts are daisychained and use CSMA/CD for acces to the medium) nor is it like a switched Ethernet. Although the data frames do look like Ethernet frames, at least typically.

By the way, the same applies to optical networks through neighborhoods. Most of these are so-called passive optical networks, or PONs. There are no Ethernet switches throuighout the neighborhood. Instead, passive optical splitters. Just like the coax plants. Which also means, upstream traffic needs to get to the headend before it can go back downstream. These PONs have used Ethernet and even ATM data frames.

Bert

Reply to
Albert Manfredi

I believe the term comes from video, not from networking. One can process audio radio signal with fairly simple filters, at least ot the RF level. At IF the bandwidth is a significant fraction of the IF carrier. For video, one much keep the passband flat over the bandwidth of the signal. A CATV amplifier will amplify all channels together, maintaining the flat passband.

For FM, the ratio of the deviation to the carrier is important, though it seems that wideband, instead of broadband, is used, in contrast to narrowband. (Broadcast NTSC TV is AM-VSB, satellite TV is often FM.)

But is it really that different from 10broad36? As well as I know it, it is an evolutionary improvement, with different, more efficient, modulation methods but otherwise not so different.

I haven't read the details recently, but I think that is the way 10broad36 works, too.

(snip)

-- glen

Reply to
glen herrmannsfeldt

Could be, I just never made that association. "Broad" could also denote broad baseband, as far as I'm concerned.

Wow, I didn't even think of that. Seems similar, although I believe the cable systems use a round robin scheme for uplink, where 10BROAD36 uses collision detection. But you're right. 10BROAD36 was developed for cable TV systems.

Bert

Reply to
Albert Manfredi

Do you happen to know what - if any - technology differences there are in the higher speed Comcast Internet services that deliver 30 Mbps and 50 Mbps download speeds to retail customers? These are branded by Comcast as Ultra and Extreme 50. A press release discusses them here:

formatting link
I have been very skeptical that these services would actually deliver anything close to the claimed speeds consistently, and I have been waiting to see some good audit data that shows what the actual throughputs are for each service, monitored over a 24x7 timeframe.

Reply to
W

It's DOCSIS 3.0, channel bonding.

formatting link
No idea how well it works personally, but others are reporting good results in the Comcast forums and on dslreports.

Reply to
Char Jackson

Virginmedia have a 50 Mbps service in the UK which is also DOCSIS 3, and are testing 100 / 200 Mbps.

formatting link

Reply to
Stephen

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.