Ray, Interesting reply. There you go, getting practical again! <s>
What you stated reminds me of how, when we're dealing with a controlled environment, we design to optimum levels with very fine granularity and resolution. And when the environment is uncontrolled, or hostile, as in the case of the extended local loop, it seems that we shotgun our way through, designing around worst-case expectations and the unknown.
I think that some interesting contrasts and parallels can be drawn between in-building systems and outdoor plant.
The TIA 568 standard and its associated annexes represent a model for in-building and campus structured cabling system design, and together these represent an example of the former, while the greater outdoor plant scenario we have been discussing here, inclusive of terminating hardware in the CO and residential/business locations, is an example of the latter.
In the case of TIA 568 we talk about the Link and the overall Channel criteria down to the number of patch panel appearances permitted, maximum distances in meters, and cross-connect hardware quality in precise and unforgiving terms. In the case of the local loop we speak in much broader terms, to the point vagueness on certain issues. I find that to be not unreasonable, for the reasons that you cite, yet, innately interesting.
In the 568 scheme of things, TDRs, as you suggest, are precisely what are used to determine distances, and certain other parameters, with NEXT being one of the most important criteria to control. While NEXT is not a concern in xDSL at this time, as George has pointed out, I can see where it may be of concern in multi-line configurations in the future, where users want to maximize at the higher 50 to 60 Mbps speeds in both directions.
It's surprising to me that there is not as much detail being placed in loop parameters and cross-connect hardware specifications (unless a model exists that I am unaware of) for the higher-speed levels of ADSL, and certainly, for VDSL.
Your points on documentation are well taken. One of the services I am very familiar with is that of automated cable management systems. I understand how difficult and tedious a task it is to manage even a controlled environment like a building or campus in this respect. And my background with both IXCs and LECs can lend ample testimony supporting your other observations and remarks.
What, if any, qualifying loop criteria do already exist for xDSL loops beyond the distance limits, insertion loss, frequency response, and the non-loading caveats? Ray? Anyone?
Frank |
