| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Hi Matt, Comments below. Hesham From: Matthew Schmitt [mailto:m.schmitt@xxxxxxxxxxxxx]
Hesham, I wasn't so much suggesting that MAC or upper layers are — or should be — in scope. Rather, when analyzing a solution, you have to take the whole system into account: you can't necessarily just
look at the PHY layer in isolation. HEB> I agree. For example, there are a number of different ways that different PHY layers adapt for plant conditions. Some of these may have minimal impact on the capacity of the PHY layer itself, and so if
you looked at the PHY in isolation, it might seem innocuous. However, it could cause a shift in how the PHY layer is able to transport MAC layer frames, resulting in a much more significant impact to the MAC layer's overall capacity/throughput than you might've
expected just from looking at the PHY in isolation. Similarly, something with a more significant impact at the PHY layer might mesh better with the MAC layer, and have less severe impacts on overall system performance than you might otherwise expect. Could all of this lead to a conclusion that we need to look at things like MPCP in order to improve efficiency? Potentially, although I think we need to understand how well we can match up various
coax PHY options to the existing MPCP before we ask that question. Thanks. Matt P.S.- I'm not sure I'm familiar with the DBA acronym in this context — could you clarify? Thanks! HEB> DBA is Dynamic Bandwidth Allocation. Various DBA algorithms have been proposed and implemented. A DBA algorithm is implemented in the OLT; it uses MPCP Report and Gate messages to build and communicate a transmission schedule to ONUs. From: Hesham ElBakoury <Hesham.ElBakoury@xxxxxxxxxx> Good input Matt! Few questions/comments below. Hesham From: Matthew Schmitt [mailto:m.schmitt@xxxxxxxxxxxxx]
Hesham, To your question about cable plant size, it is likely most of the above, rather than any one particular metric, as several of those can impact the performance of the PHY layer (or even the MAC
layer and associated sub-layers). HEB> I think MAC and upper layers are out-of-scope in EPoC based on our CFI discussions. The MAC is Ethernet MAC. There might be some new OAM messages. Do you think MPCP needs to change
to support the characteristics of the HFC plants or MPCP should stay the same and DBA may need to change ? (as you know DBA is out of scope of IEEE 802.3 standards). Although Mark and Howard may have had something more specific in mind with that particular question. For example, for a given node, we'll need to define (or select) ranges or approximations for things like homes passed, length of coax segments, number of amplifiers in cascade, arrangement of amplifiers
(for funneling effects), fiber lengths, number of value of taps, tap performance, etc. This will likely be critical in evaluating different proposals, and understanding how they'll work under various conditions. We'll probably also need to define service
group size, and how many nodes are covered in a given service group (assuming the architecture supports multiple nodes in a service group). HEB> I would appreciate if Mark/Howard can confirm that these are the characteristics of the plant, amplifiers, taps that they are looking for or we need to take into account other characteristics such as amplifier frequency response, gain bandwidth product, distortion reduction, …. etc Note that Howard and Mark did call out a number of the things I noted above separately from "plant size" -- which is why I suspect that they may have had something more specific in mind — but IMHO
it all kind of lumps together in terms of characterizing the plant conditions under which a proposal will likely need to be evaluated. Thanks. Matt From: Hesham ElBakoury <Hesham.ElBakoury@xxxxxxxxxx> Hi Howard, I have few questions regarding the Operator requirements (pertaining to physical layer): Service coexistence issues and criteria Services to be provided over EPoC HEB> Do you mean residential and business services or video, IPTV, VoIP, data services ? and how they will evolve over time Business vs residential services
(will they exist on the same network?) Asymmetry vs. symmetry Existing cable plant characteristics Architectures
(Node+0 "passive", Node+N [N=1-?], Complete HFC, MxU)
Amplifier characteristics and considerations that will effect the PHY Cable & passives characteristics Typical size of cable plant HEB> I am not sure how you measure the size ? (number of home passed, size of coverage area, number of nodes, … etc). Subscribers passed Number and size of taps Changes to cable plant characteristics over time
(e.g. passive and active element changes, any use of bypasses?) Spectral allocation and how it changes over time Which frequencies are amplified and which are passive ` What spectrum will be allocated for EPoC initially,
and how will that change over time Regional differences for changes? Functional Assumptions and Impairments DOCSIS 3.0 has already characterized in CM-SP-PHYv3.0-I05-070803, Chapter 5, and in CM-SP-DRFI-I12-111117, Chapter 5, for "in amplified" regions of
cable. For both "in amplified" and "passive" EPoC considerations, are
there any additonal functional assumptions and impairments that need to
be considered for up to 1Gbps and higher operation?
How will these change over time? Are there regional differences; e.g. China, Europe? Number of subscribers per network, take rate Minimum required channel data rate Maximum desired channel data rate Thanks Hesham -----Original Message----- The first meeting of the IEEE 802.3 EPON Protocol over Coax (EPoC) PHY Study Group will be held January 24th and 25th in Newport Beach, CA, hosted by the Ethernet Alliance. Please see my previous message for
links to meeting logistical details. In preparation for the meeting, Mark Laubach and I put together a list of topics that can help us prepare for a successful study group meeting, and these are listed in the attached file. The first set of topics deals with the IEEE 802 standards development process. I have all of the material I need for this section, and I am sure that I will be able to enlist the help of some of our experienced
hands to deliver it. The second set of topics deals with operator requirements, and this is where I would like to make an appeal for contributions. We may not get contributions that address all of the topics listed, and there may
not even be universal agreement that the topics are relevant, but I think that these are areas that must be explored in the study group if we are to do a proper job of defining the scope of a new project, and judging it against the "5 Criteria". People wishing to make a contribution to the study group should review the "Procedure for Presenters" information that can be found here: http://www.ieee802.org/3/epoc/public/presentproc.html Since this is a new email reflector, I would also like to make people aware of the IEEE 802.3 Working Group email reflector policy that can be found here: http://www.ieee802.org/3/reflector_policy.html I will welcome your comments and questions. Howard Frazier Broadcom Corporation ________________________________________________________________________ To unsubscribe from the STDS-802-3-EPOC list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-EPOC&A=1 <="" p="">
|