Quote:
Originally Posted by Ignitionnet
...
Congestion is caused by having too many modems / too much demand on a particular CMTS port. The optical node sizes are very relevant to this as each optical nodal area is a 'broadcast domain' - within a single node the only option is to run more channels.
|
Which then begs many more increasingly complex questions that attempt to speculate on why VM have some perversely protracted fix times in some areas?
For example:
1. Does Cisco licensing still depend on number of "channels" used per CMTS?
2. The number of "usuable" channels will depend on quality of RF spectrum. This must largely depend on combination of amount of domestic noise injected upstream and also total number of cm's to avoid laser clipping? Originally 10+ yrs ago the DOCSIS RFI spec ,number of street cabs, optical nodes would have been designed on basis of HPPN (homes passed per node) and an estimated "takeup" of say 10/20% subscribers? In some older Victorian/Edwardian built towns/city's (particularly those with higher student populations) 3 story town houses and similar have been progressively converted to self contained flats over the intervening period which appears in some cases to have resulted to congested tap boards and frequent "re-splitting" of original house drops. This must have some deleterious affect on both "noise" and upstream power by effectively increasing the HPPN and/or subscriber "takeup" per node?
3. If current RF spectrum is limited then the traditional approach is to re-segment. Presumably since most (or hopefully all) of the legacy FP lasers are replaced with WDM does this mean just splitting a coax trunk or would this still increase the risk of laser clipping?
3a. Since power for trunk amps/line extenders is injected at optical node this could imply co-requisite upgrades to power supply is also required in some cases from utility company? Interestingly as an aside note, I thought UK used 48v DC for trunk/amps,line extenders not AC?
3b. If dynamic load balancing is also restricted to within same PHY RF segment does this mean that either the FO has to be "uncombined" at CMTS or separate FO blown down to an additional optical node?
4. Presumably 3b is also dependent on the amount of "dark fibre" available again which could vary by area?. If not then "ducting" capacity/limitations could very well imply substantial civil engineering work?
5.Would upgrading to high density linecards increase either the capacity of CM's per channel,the number of channels/upstream ports or both? Again as well as more power efficient I thought these reduced the need to physically re-segment?