[ndnSIM] [EXT]Re: Self-learning strategy not working properly

[Teng Liang]

Dear Andrew,

In NFD, the self-learning strategy relies on the RIB manager to manage FIB
updates, and the RIB manager is running on a separate thread. In ndnSIM,
the RIB manager/thread seems not to be imported with NFD, hence the
self-learning strategy cannot work. One possible solution is to modify the
self-learning strategy to update FIB directly instead of relying on the RIB
manager in ndnSIM (by replacing code snippets start with *runOnRibIoService*
).

Best regards,
Teng

On Wed, Sep 2, 2020 at 12:22 PM Andrew S via ndnSIM <
ndnsim at lists.cs.ucla.edu> wrote:

> *External Email*
> After all my attempts it still fails like shown before.
> The main problem is that it cannot correctly send data back - in most
> cases data is sent back only when it hits cache (and for a single hop, as
> things repeat on the next one).
>
> If anyone had any experience with Self-Learning strategy, please, let me
> know.
>
> ср, 26 авг. 2020 г. в 20:00, Andrew S <forelf1 at gmail.com>:
>
>> Hello,
>> I'm currently working on implementing custom forwarding strategy and
>> would like to compare it to the existing ones.
>>
>> I could not find any examples of Self Learning strategy, thus I just
>> modified the ndn-grid scenario to use self-learning strategy instead of
>> best-route one. I as well removed the GlobalRoutingHelper (as far as I
>> understand Self strategy does not need it).
>> The outcome of the simulation is not what I expected:
>>
>> Time Node AppId SeqNo Type DelayS DelayUS RetxCount HopCount
>> 4.11533 0 0 1 LastDelay 0.115328 115328 1 4
>> 4.11533 0 0 1 FullDelay 3.11533 3.11533e+06 2 4
>> 7.0865 0 0 0 LastDelay 0.086496 86496 1 4
>> 7.0865 0 0 0 FullDelay 7.0865 7.0865e+06 3 4
>> 12.1153 0 0 2 LastDelay 0.115328 115328 1 4
>> 12.1153 0 0 2 FullDelay 9.11533 9.11533e+06 2 4
>>
>> So just three satisfied interests over 20 seconds with CBR of 1 interest
>> per second.
>>
>> This is the default latest ndnSIM with no modifications except for the
>> ones I described above.
>>
>> Could somebody clarify to me if this is the expected way of how
>> self-learning strategy works? Or am I missing something?
>>
>> Thanks in advance for any help!
>>
>> --
>> Andrew S.
>>
>
>
> --
> Andrew S.
> _______________________________________________
> ndnSIM mailing list
> ndnSIM at lists.cs.ucla.edu
> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndnsim/attachments/20200922/fd1b7e5d/attachment.html>