[ndnSIM] Flow control and congestion control in ndnsim

[Lixia Zhang]

> On Mar 24, 2016, at 3:32 AM, Ahmed Sadek <don1559 at gmail.com> wrote:
> 
> Thank you all for your replies, I really appreciate it.
> 
> I have googled and found some papers but was interested to find out if there was implementation for those ideas in NDN or ndnSIM.

no.
but people can help implement in ndnSIM

> Another question, NDN is pull based so flow control is controlled by the receiver but what will happen if the link between source and destination change bandwidth capacity (2 mega -> 1 mega ) or have data more than it's capacity (3 mega data request over 2 mega link ) , does data packets gets dropped and interest re transmitted ? so what is the expected behavior in NDN and ndnSIM ?

"the expected behavior in NDN and ndnSIM" sounds a bit ambiguous.
Lets separate out what can be done from what has been implemented.

NDN's 2-way interest-data exchanges form a feedback loop, enabling intelligent forwarding.
in your above example: the router knows how fast it can pull in data before the BW change, and can notice the failure after the change. The router can try alternative path if one exists, or send a NACK back to the prev hop (wonder if you had taken a look of the paper I pointed to below?  paper http://named-data.net/wp-content/uploads/comcom-stateful-forwarding.pdf <http://named-data.net/wp-content/uploads/comcom-stateful-forwarding.pdf>)

> On Wed, Mar 23, 2016 at 7:59 PM, Lixia Zhang <lixia at cs.ucla.edu <mailto:lixia at cs.ucla.edu>> wrote:
> 
>> On Mar 23, 2016, at 11:25 AM, Spyridon (Spyros) Mastorakis <mastorakis at CS.UCLA.EDU <mailto:mastorakis at cs.ucla.edu>> wrote:
>> 
>> Hi Ahmed,
>> 
>> this is a quite open-ended question about the NDN architecture in general. Currently, there is no way to perform congestion control in NDN.
> 
> this seems some English expression issue -- there are plenty of ways to perform congestion control (e.g. see some ideas from this paper http://named-data.net/wp-content/uploads/comcom-stateful-forwarding.pdf <http://named-data.net/wp-content/uploads/comcom-stateful-forwarding.pdf>)
> 
> I believe Spyros meant that today's NFD implementation has not adopted any specific solutions, hence he has not put a congestion control into ndnSIM. 
> 
> Klaus Schneider is working on CC as Spyros already mentioned; a quick google search on "ndn congestion control" also returns lots pointers to the papers on the topic
> 
> to the original question:
> 
>>> On Mar 21, 2016, at 3:44 AM, Ahmed Sadek <don1559 at gmail.com <mailto:don1559 at gmail.com>> wrote:
>>> 
>>> Dear All,
>>> 
>>> I was wondering if someone can point me to information on the Flow Control and Congestion Control mechanism used in ndnsim and is it the same as in NDN ?
>>> 
>>> Also, if the client keep sending interests on different interfaces and not receive response then what is the expected behavior ?
> 
> the question is: what's the cause, right?
> e.g.
> - if the node is not connected, that's a connect problem
> - if there is local connectivity but the network is partitioned, that's a slightly different problem
> - if the interest the data does not exist, ideally one should receive an app level NACK as explained in http://named-data.net/wp-content/uploads/2016/01/consumer_producer_communication.pdf <http://named-data.net/wp-content/uploads/2016/01/consumer_producer_communication.pdf> 
> - if the named producer does not exists, then it is up to how your strategy works (could generate a "NO ROUTE" NACK)
> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndnsim/attachments/20160324/f08033a0/attachment.html>