Re: [802.3_EPOC] Study Group Questions
Hi Mr. Yao:
Thanks for your comments, I see your points. The situation in China may be different compare with here in US. North America has large active HFC plants with N+x, it is very expensive to move to N+0. We need a solution scale from N+x to N+0 over the years. EPON MAC is designed with FDD in optical domain; so FDD to coax is a minimum impact maximum protocol reuse extension. Beside, it provides backward compatibility with EPON.
TDD, on the other hand, is a RF spectra efficient technology; it makes sense for native Ethernet over cable since EOC is a new MAC. In EOC we have all the freedom to design an optimum MAC for cable. The philosophy of EPOC is reusing EPON MAC to schedule on both optical and cable sections; we are more confined on what we can do without affecting the backward compatibility and simplicity. Since EPON MAC is designed for fiber, it may not optimum for coaxial cable considering the fundamental cable characteristics, such as echo delay, etc. There is a tradeoff.
I believe that we may do better on Native Ethernet over cable with new protocols designed for the coaxial cable; we can carry on this conversation outside the WG.
Happy New Year and Chinese New year,
Eugene
-----Original Message-----
From: 姚永Gmail [mailto:yy0412@xxxxxxxxx]
Sent: Wednesday, January 04, 2012 4:45 PM
To: Dai, Eugene (CCI-Atlanta); STDS-802-3-EPOC@xxxxxxxxxxxxxxxxx
Cc: wangzhao; 高晓俊; rwg@xxxxxxxxxx; 秦 yanlong
Subject: Re: [802.3_EPOC] Study Group Questions
Eugene Dai,
我正希望了解Cox有一部分HFC网络改造为3GHz是如何进行的，是有源的还是无源的？有源的话，频谱是如何分割的？能适应EPOC应用吗？放大器如何设置？把原有的1GHz以下同轴网络升级为3GHz同轴电缆的放大节距是否需要改变？整个改造工程和光纤延伸相比造价有没有优势？不光要考虑一次性改造投资，还要考虑长期维护成本。美国人力成本比中国高很多，因此情况可能会不一样。在中国，这种有源同轴改造是不值得的——光纤比面条便宜。如果考虑长期维护和电力消耗更是如此。因此中国目前大规模进行的都是光纤到楼的改造（同轴无源化），已经改造的区域估计达到20~30%，而且这个比例在迅速增加。
TDD并非只适用于家庭网，只要N+0，接入同样适用。4G移动通信就是TDD、FDD都支持的。
当静态分配上下行带宽（但是可以随时改变配置）时，TDD和FDD没有本质区别，包括对EPON MAC的适配。目前中国市场上的ECAN就是采用EPON MAC的一种技术，上下行带宽是静态分配的；DECO是另一种采用EPON MAC的技术，而且上下行带宽是动态分配的。
FDD有许多优点，但对于EPOC这种需要大带宽频谱的应用而言，频谱的规划非常困难，频谱适应性不如TDD。因此我们主张FDD和TDD都需要支持——FDD支持N+n，频谱规划没有问题的地方，TDD支持N+0，需要灵活分配频谱的地方。
多交流、多沟通！
祝好！
姚永
----- Original Message -----
From: "Eugene Dai" <Eugene.Dai@xxxxxxx>
To: <STDS-802-3-EPOC@xxxxxxxxxxxxxxxxx>
Sent: Wednesday, January 04, 2012 11:46 PM
Subject: Re: [802.3_EPOC] Study Group Questions
> From the HFC evolution point of view, moving from N+m to N+n, where n<m, will be a progressive process with extended time lines; at the end of this process whether it will be N+0 is still a question as far as economy is concerned. The strength of Ethernet/EPON over Cable is the potential to provide more scalable upstream and downstream data services over HFC plants without the expensive N+0 upgrade. Therefore, passive cable plant should NOT be an assumption for EPOC project.
>
> TDD – Time Division Duplex was introduces during the early evolution process of HFC under constrains of limited RF spectra availability in the analog/QAM video dominated cable plants. It is a temporally solution to certain degree. Some of the TDD based EOC solutions were adopted from home networking standards such as Homeplug and MoCA that are not scale well in access network.
>
> From the technology point of view, TDD is half duplex; Ethernet has moved away from it long time ago. As has been pointed out, EPON does not support half duplex mode. In order to make TDD work in the application we targeted for, either EPON MAC need to be changed substantially or a cable MAC need to be introduced; either way backward compatibility with EPON will be a problem.
>
> Regards,
>
> Eugene
>
> -----Original Message-----
> From: Chris (Yanbin) Huang [mailto:chrish@xxxxxxxxxxxx]
> Sent: Wednesday, January 04, 2012 12:08 AM
> To: STDS-802-3-EPOC@xxxxxxxxxxxxxxxxx
> Subject: Re: [802.3_EPOC] Study Group Questions
>
> I would like to translate Mr.Yao following email for your reference.
>
> "Is it good for EPOC to fit for some portion of Markets, or all Markets? FDD technology implementation is simplier, but its spectrum allocation plan is very difficult, especially for the whole global, is very difficult to be unified. It is different from DOCSIS: one Downstream channel of DOCSIS only needs 8MHz and narrower spectrum. If FDD mode is used, EPOC should take 10G EPOC plan, otherwise, the following upgrade cannot be backforward-compatible, but 10G system spectrum allocation is very difficult to plan. For neighboring channel interference, EPOC system also is completely different from DOCSIS: DOCSIS is based on multi-channels, so FDD has big advantage, but EPOC is based on single channel (it is very difficult to find the spectrum meeting multi-channels). "4G" tech just is to support both TDD and FDD. FDD is relatively eaiser to support 'active relay', but I doubt whether it is worthy, considering spectrum allocation plan difficulty and coax network reconstruction !
in!
> vestment: It is better to extend fiber reach and remove amplifiers, than to reconstruct/upgrade coax network.
>
> Chris (Yanbin) Huang
> ________________________________________
> From: 姚永Gmail [yy0412@xxxxxxxxx]
> Sent: Tuesday, January 03, 2012 3:21 PM
> To: STDS-802-3-EPOC@xxxxxxxxxxxxxxxxx
> Subject: Re: [802.3_EPOC] Study Group Questions
>
> 是只适应一部分市场好还是适应所有市场好？
> FDD技术实现比较简单，但频谱规划很困难，特别是针对全球，很难统一。这和DOCSIS不同——一个下行信道只需要8MHz及以下频谱，如果采用FDD方式，应该按照10GEPOC规划，否则后续升级无法后向兼容，而10G系统频谱很难规划；对于邻信道干扰，在EPOC系统也和DOCSIS完全不同——DOCSIS是多信道应用，FDD有很大好处，但EPOC是单信道应用（很难找到满足多信道的频谱）。4G就是同时支持TDD和FDD的。FDD比较容易支持有源中继，但考虑到频谱规划的难度和改造的投资，我很怀疑是否值得——与其改造同轴网，还不如光纤延伸，取消放大器。
> 姚永
> ----- Original Message -----
> From: "Matthew Schmitt" <m.schmitt@xxxxxxxxxxxxx>
> To: <STDS-802-3-EPOC@xxxxxxxxxxxxxxxxx>
> Sent: Wednesday, January 04, 2012 6:46 AM
> Subject: Re: [802.3_EPOC] Study Group Questions
>
>
>> Alex,
>>
>> In response to your comment below to me...
>>
>> I'm not sure I agree that a TDD system inherently provides more spectral
>> flexibility than an FDD system in a passive coax environment. Depending
>> on the PHY layer, you can do quite a bit of steering regardless of FDD vs.
>> TDD. I will agree that a TDD solution can provide advantages in terms of
>> the relative allocation of bandwidth to upstream vs. downstream operation,
>> although only with a MAC designed to support such features.
>>
>> That said, I'm more than open to being proven wrong, and so I will
>> definitely look forward to your presentation to highlight why you believe
>> TDD had advantages over FDD.
>>
>> BTW, I also tend to agree with others on this thread that it's MUCH
>> preferred if we can develop a single solution. Without that, you end up
>> with a fragmented market, and it's much harder to achieve the same
>> economies of scale that you could have with a single, unified solution.
>> If it's simply not possible to come up with a single unified solution, so
>> be it; but I think that should be our goal if at all possible.
>>
>> Thanks.
>>
>> Matt
>>
>>
>>
>> On 12/15/11 7:48 PM, "Liu, Alex" <alexliu@xxxxxxxxxxxxxxxx> wrote:
>>
>>>Ed,
>>>
>>>Let me reiterate some of the finer points of my previous missive: (1) TDD
>>>systems have a wide footprint in China in part due to their spectral
>>>convenience, (2) TDD should be an optional profile in addition to and not
>>>in place of FDD. I do believe that market enthusiasm for, and acceptance
>>>of, this standard and its ensuing products should be our guiding
>>>principle, and not doctrinal orthodoxy within a standards framework. If
>>>things weren't so, Ethernet would have never abandoned CSMA/CD for
>>>first-mile applications.
>>>
>>>More specifically, it is clear that an FDD RF system that directly maps
>>>to the dedicated wavelengths in fiber and thus the EPON protocol is most
>>>appropriate for the N. American MSO environment. I would like to raise
>>>the possibility that this is not necessarily true for China. Passive
>>>cable plant coupled with haphazard spectrum planning makes for an
>>>inviting TDD target. There are then the orthodoxies emanating from the
>>>Chinese side. If we are serious about targeting the China market, I
>>>suggest we consider SARFT's input.
>>>
>>>@Matt: TDD's ability to operate in unpaired spectrum makes "lively"
>>>spectrum plans possible in China. Perhaps this is improperly termed
>>>"coexistence with" and is better called "steering around" existing TV and
>>>data systems. This additional degree of freedom may perhaps be attractive
>>>to N. American operators as well.
>>>
>>>@Mark: we *are* working toward a single standard. Transparent EPON
>>>protocol operation over coax is the goal and FDD RF operation should be
>>>the mandatory supported mode. Employing the modern PHY proposals being
>>>developed in an optional TDD mode should not detract from this stated
>>>goal. LTE offers an instructive precedent.
>>>
>>>Alex
>>>
>>
>> ________________________________________________________________________
>>
>> 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
>
> ________________________________________________________________________
>
> 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
>
> ________________________________________________________________________
>
> 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
>
> ________________________________________________________________________
>
> 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
________________________________________________________________________
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