[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [WG-BB] Study Report: CATV assignment guidline
Maemura-san,
Thank you for your contribution!
At 13:34 02/02/04, MAEMURA Akinori wrote:
>Colleages,
>
>WG-BB report was presented in the last Policy SIG in Taiwan.
>We JPNIC IPwg reported it in JPNIC Open Policy Meeting but
>there some SIers and CATV operators presented their opinion
>around default bootstrap assignment size.
>
>JPNIC IPwg had a meeting with concerned people and studied
>this issue. The Following is the study report.
>
>As I asked Arano-san with a hat of chair of Policy SIG, I'd
>like to present something based on this discussion, after
>discussing in this wg-bb.
Generally, it is OK. However, it should depend on discussion
which will happen in this ML.
>Summary:
>------------------------------------------------------------
> a) Default assignment size should be considered upon
> "number of downstream port" and
> "capacity specification" by the hardware,
> instead of simply number of CMTS.
Could you give us more concrete conditions and
number (/24, /23 or whatever)?
> b) In case of one 30Mbps type cable, /24 will typically
> last three months. Considering it can connect 2,000
> users for example, it is never harmful if default
> size for this is /23 which lasts six months.
>------------------------------------------------------------
Do you mean one 30Mbps type cable is automatically given /23?
I can't see any condition about "number of downstream ports"....
Regards,
Takashi Arano
>Discussions:
>
> 1) CMTS itself has various port density and capacity.
> some can connect multiple cables and there are two
> types of specifications which allow 10Mbps and 30Mbps.
> The difference of the capacity will be clearly
> distinguished by product name.
>
>
> 2) some CMTS has a configuration with different numbers
> of upstream/downstream ports. This is because down-
> stream cable can be divided into multiple with no
> significant signal loss but cable aggregation at up-
> stream cable will increase noise level so such a CMTS
> is designed to connect upsteam cables with no cable
> aggregators.
>
> Here the number of downsteam cables is the most
> appropriate index for deciding default assignment
> size if capacity specification is taken account into.
>
>
> 3) Maximum number of customers for a cable is suggested
> for each hardware for example 2,000 but the practical
> number will vary according to the design policy of
> each operator.
>
>
> 4) The proposed default assignment size that is a /24
> for one CMTS is meets design condition of
> a) 30Mbps type CMTS connecting one downstream
> b) Multiplexing late 1000%
> c) 1Mbps throughput per user
>
>
> 5) Statistics of Cable Internet in Japan says average
> distributed capacity for one Cable Internet user is
> 5Kbps to 12Kbps. It corresponds ten times as many
> customers as scenario in 4) while it is just an
> avarage and it depends on service level design of
> each operator.
>
>
> 6) Customer demand in the bootstrap phase of CATV
> depends on its circumstances like existing xDSL
> services in the area, IP awareness or someting like
> these. But /24 is likely to be a size that it lasts
> three months in bootstrap in a typical case.
>
>
> 7) In case of cisco hardware, multiple secondary
> addresses can be assigned to single cable interface
> so that Operator can easily add another assignment
> after it is implemented, however some hardwares do
> not allow this and single subnet must be assigned
> there.
>
>------------------------------------------------------------
>
>
>
>Regards,
>Akinori
>-
>- This list (wg-bb) is handled by majordomo@lists.apnic.net
-
- This list (wg-bb) is handled by majordomo@lists.apnic.net
