Hi Tom,

I would like to assist on that. I'm finishing writing my Master Degree
Thesis comparing DS-Lite, 464XLAT, MAP-E and MAP-T performance. I got of
one CGN + encapsulation (DS-Lite), one CGN + dual translation (464XLAT),
one A+P + encapsulation (MAP-E) and one A+P + dual translation (MAP-T).

At the moment I have collected performance (throughput, jitter, latency,
packet loss and overhead) and user experience (Internet access, Torrent,
Skype etc).

I used a small scale network with ISC's DS-Lite implementation and
CERNET's MAP-E and MAP-T implementation. For the 464XLAT I had to use
Tayga, as the current solution being used at T-Mobile is Android only
and Tayga was the only way I found to implement 464XLAT without using
Android and allowing me to have the same network for all the tests.

I was already planning to write a summary of my results as a draft when
I had finished everything. Your proposed scope is larger than mine as
you are proposing to analyze more techniques and do planning /
management considerations, but I believe I can assist you.

Edwin Cordeiro

Em 06/03/14 17:36, Tom Taylor escreveu:
> Comments and additions are invited. If there is no enthusiasm for the
> project I'll drop the subject.
>
> Scope
> -----
>
> This would cover the various Softwires 4 across 6 approaches: DS-Lite
> (with GI DS-Lite a special case), MAP-E, LW4o6, MAP-T, 4rd.
>
> Method
> ------
>
> In summary, the analysis would proceed using two or three basic
> scenarios. The operation of each approach within each scenario would
> be analyzed by walking through how it would work for various aspects.
> All approaches would be evaluated against the same set of aspects
> (more details below). This preliminary analysis would be completed for
> all approaches before any conclusions are drawn.
>
> Going forward, differences between the different approaches would be
> identified. These would form the core of the I-D in its finished form.
> The text reflecting the initial analysis would be discarded if it
> wasn't interesting or moved to appendices if it was.
>
> Scenarios
> ---------
>
> At a minimum, we should use two scenarios: an established fixed
> broadband network, and an established LTE network.  Suggestions
> welcome here if this list should be expanded or the scenarios
> described more elaborately.
>
> Aspects of the Walk-Through
> ---------------------------
>
> I see the following aspects as a preliminary list. Again, comments are
> invited.
>
> Numbering Plans
> - IPv4
> - IPv6
>
> Routing
>
> Operations
> - bringing a border router into service
> - adding a subscriber
> - moving a subscriber
> - logging
> - lawful surveillance
>
> Scalability
> - growth in number of subscribers
> - growth in traffic
>
> Fault Management
> - resilience to network failures
> - traceability of subscriber issues
>
> Network Attachment
>
> Data Plane Issues
> - fragmentation
> - multicast
> - check sums
>
> QoS
> - delay
> - jitter
>
>  Tom Taylor
>
> _______________________________________________
> Softwires mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/softwires

_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires

Reply via email to