
how-to block ads
|
|
Share Topic  |
 |
|
|
 DMS1 join:2005-04-06 Carrollton, TX | reply to patcat88
Re: coax Ethernet from the node?????? said by patcat88:If your running ethernet over coax for the last mile, why dont they deploy ethernet over coax from a fiber node for the last mile, much higher speeds, less overhead, less D2A, A2A, A2C conversions. The DOCSIS 3 to Coax Ethernet convertor isnt going to be a cheap piece of equipment (its active electronics), and it eats line power, why dont they just do Coax Ethernet to the home stright from the nice fat fiber node? Presumably because that would involve a cabling infrastructure upgrade too. A typical HFC network consists of a node that drives a hundred or more drops via a series of amplifiers and multi-port taps. I guess that what NTL are planning to do is place a multi-user DOCSIS 3 modem in each tap (which from what I remember have many more ports than the typical tap used in the US), meaning that each user then has a dedicated coax link from the modem to their house. With your plan, multiple Ethernet-over-coax links would need to be carried from the node to the taps, and this may be beyond the capabilities of the existing coax. | |  patcat88 join:2002-04-05 Jamaica, NY kudos:1 | said by DMS1:Presumably because that would involve a cabling infrastructure upgrade too. A typical HFC network consists of a node that drives a hundred or more drops via a series of amplifiers and multi-port taps. I guess that what NTL are planning to do is place a multi-user DOCSIS 3 modem in each tap (which from what I remember have many more ports than the typical tap used in the US), meaning that each user then has a dedicated coax link from the modem to their house. With your plan, multiple Ethernet-over-coax links would need to be carried from the node to the taps, and this may be beyond the capabilities of the existing coax. Isnt the speed increase nullified by sharing a individual modem? | |  DMS1 join:2005-04-06 Carrollton, TX | said by patcat88:Isnt the speed increase nullified by sharing a individual modem? I suspect that the "device" is actually a multiple-modem unit. In other words, it offers the same functionality as multiple individual modems but in one physical device, allowing the sharing of some circuitry such as the node-side interface, power supply, control function etc. This sharing would ensure that an N-port modem was considerably cheaper than 'N' discrete modems. | | |
|
|