[ndnSIM] Interest is not re-broadcasted using the same wireless face in ndnSIM

Jason Christian jason.christian.92 at gmail.com
Wed Oct 4 10:28:09 PDT 2017


Hi,

thank you for the reply. I thought that the interest packet can only be
forwarded on non-local face, since the consumer itself has a non-local face
and the rest of the nodes (Producers) only have local face.

I have tried another alternatives, which is to change the propagation model
of the channel to ThreeLogDistance with Nakagami (earlier before, I used
RangePropagationLossModel) and the nodes actually forward the Interest
packet to the neighboring nodes. Is it probably the problem with the
RangeModel?. I have also calculated the node's position to each other and
make sure there will always be nodes within range of a node, so the
RangePropagationLossModel should also work

Another thing. Since I have a custom strategy (this custom strategy also
takes into acct node's position, much like geocasting) and a custom app,
where should I put the Interest forwarding code? Should I:
- Forward them using Strategy::sendInterest?
- or just copy the Interest on the custom app and send it?

Sorry for the seemingly silly questions

Best Regards,
Jason

On Wed, Oct 4, 2017 at 6:54 PM, Spyridon (Spyros) Mastorakis <
mastorakis at cs.ucla.edu> wrote:

> Hi,
>
> if you think that the provided face abstraction is not sufficient for your
> simulations, then feel free to add a custom NetDevice Face implementation
> and use that for your simulations.
>
> I do not fully understand your previous email and the problem here. Some
> comments off the top of my head though:
>
> 1) a process (i.e., an application running on a host) should communicate
> with a local face with the local forwarder. The difference between a local
> and a non-local face has to do only with whether Interests with the
> /localhost and /localhop prefix can be forwarded through them.
> 2) Interests with any other prefix can be forwarded through local and
> non-local faces.
>
> Please take a look at the NFD developer’s guide for all those definitions:
>
> https://named-data.net/wp-content/uploads/2016/10/ndn-
> 0021-7-nfd-developer-guide.pdf
>
> Thanks,
>
> Spyridon (Spyros) Mastorakis
> Personal Website: http://cs.ucla.edu/~mastorakis/
> Internet Research Laboratory
> Computer Science Department
> UCLA
>
> On Oct 3, 2017, at 10:06 AM, Jason Christian <jason.christian.92 at gmail.com>
> wrote:
>
> Hi, thank you for the reply.
>
> I have traced back from the Strategy::sendInterest() to the
> generic-link-service implementation and compare it between the mobile
> package's implementation and the one from my machine and I cannot find any
> differences. Can you also point out which particular commits I should pay
> attention to?
>
> Another thing I have seen from my custom strategy:
> I checked the method violatesScope and wouldViolateScope from https:/
> /github.com/4th-ndn-hackathon/ndnSIM-Mobile-Simulation-Package/blob/
> 5714833ccce551cf4cd128967a21283bc3ed4319/NFD/daemon/fw/algorithm.cpp#L59,
> and it will return true if the face used is a local face. In the custom
> strategy extending from Strategy class, the outFace from the nextHopList is
> always a local face. Perhaps is this the reason why the Interest packet is
> not forwarded?
>
> inFace is the face from the argument from afterReceiveInterest method
> outFace is the face from the iteration of fib::NextHopList::const_iterator,
> similar to the Broadcast strategy (I based mine on that)
>
> Note that the outFace on the consumer (a.k.a the first node) has
> non-local face, and thus is able to forward Interest packets to nodes
> within range. The outFace from the Producers (a.k.a 2nd - last nodes)
> always have local face (from the debug msg). If so, how can I obtain the
> non-local face of producer nodes?
>
> Best Regards,
> Jason
>
>
> On Tue, Oct 3, 2017 at 6:03 PM, Spyridon (Spyros) Mastorakis <
> mastorakis at cs.ucla.edu> wrote:
>
>> Hi,
>>
>> we have done something similar for a mobile simulation package for ndnSIM
>> (it was an NDN hackathon project). We have addressed those issues there, so
>> please take a look at the commits:
>>
>> https://github.com/4th-ndn-hackathon/ndnSIM-Mobile-Simulation-Package
>>
>> Thanks,
>>
>> Spyridon (Spyros) Mastorakis
>> Personal Website: http://cs.ucla.edu/~mastorakis/
>> Internet Research Laboratory
>> Computer Science Department
>> UCLA
>>
>> On Oct 3, 2017, at 4:17 AM, Jason Christian <jason.christian.92 at gmail.com>
>> wrote:
>>
>> Hi,
>>
>> I am currently doing a scenario using ndnSIM as the simulation tool. The
>> scenario includes:
>> - several nodes with mobility installed
>> - Wifi with 802.11a standard, wifiChannel with
>> ns3::RangePropagationLossModel and limited maxRange, and with type adhoc
>>
>> Basically, the consumer nodes will send the Interest through the wireless
>> face to nodes within the MaxRange. This works perfectly on first
>> transmission, but the Interest is not forwarded to other nodes which is in
>> range with the receiving nodes (In other words, it is not forwarded, there
>> is only 1 hop)
>>
>> I have already looked into https://github.com/cawka/
>> ndnSIM-nom-rapid-car2car/tree/master/extensions and the implementation
>> is similar to mine. The problem is, it is using ndnSIM 1.0 and the custom
>> strategy methods have been changed on ndnSIM 2.0+, which I am using right
>> now. I've also read somewhere that nodes do not actually re-broadcast
>> Interest and Data packets on the same face it received the packet from.
>>
>> I am new to ndnSIM (Have just spent around 3 days working around custom
>> scenario), so can you please guide on possible solutions of this problem?
>> - Should I create a custom face for this scenario? If so, can u point me
>> on how to do it?
>>
>> Thank you,
>> Jason
>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndnsim/attachments/20171004/888fdcaf/attachment-0001.html>


More information about the ndnSIM mailing list