[ndnSIM] Data delivery problem in NDN based Mobile Ad-hoc network (V2V scenario)

Sonia zohanoor67 at gmail.com
Thu Nov 19 00:10:03 PST 2020


Thank you for the comments. I checked the PIT entry and there are no PIT
entries in the intermediate nodes and that is why it considers the data
unsolicited and gives it to the unsolicited data pipeline.
I am trying to figure out why there is no PIT entry in the intermediate
nodes since they successfully forward all the Interest packets to the next
wireless nodes. But I couldn't find the reason yet.
I did not change anything in the source code of forwarder.cc So, I don't
know why this problem exists.
Any comment would be appreciated.


On Mon, Nov 2, 2020 at 5:21 PM DJEMAA Adel <a_djemaa at esi.dz> wrote:

> Hi,
> I haven't currently access to my simulation environment (ndnSIM), and I am
> not familiar with Wifi NetDevice of NS3.
> I have rather manipulated the LrWpan NetDevice in ndnSIM.
> For your first setting, when the producer and the consumer are
> directly connected (1 hop), you noticed an important packet loss, maybe due
> to configurations of the Wifi channel, so I think that you should test many
> configurations with different values (loss propagation model, coordinates
> of nodes, and so on) to understand the behavior of your simulations.
> For your second concern, as you have explained, the Interest packet gets
> to the producer via multihop communication, but the Data packet doesn't
> reach the consumer on the reverse path, where intermediate routers reject
> that Data packet.
> The only reason to consider Data Packet as unsolicited, according to NFD
> source code, is that there is no corresponding entry in PIT.  So
> I advise you to debug using NFD Logs, and print the PIT entry of every
> intermediate node to analyze exactly what is happening, such as the
> following command:
> *NS_LOG=nfd.Forwarder ./waf --run=your_simulation_scenario*
> Another issue, maybe is the forwarding strategy you're using, for example:
> *ndn::StrategyChoiceHelper::InstallAll("/prefix", "/localhost/nfd/strategy/best-route");*
> Or
> *ndn::StrategyChoiceHelper::InstallAll("/prefix", "/localhost/nfd/strategy/multicast");*
> In my case, I have used the second strategy, so you should test different strategies.
> Hope that helps you.
> Yours,
> Le lun. 2 nov. 2020 à 08:07, Sonia <zohanoor67 at gmail.com> a écrit :
>> Hi Adel
>> Thank you for your informative response. You are right. It was my
>> mistake, I debug my simulation again and there is only one interface 256
>> for receiving and transmitting packets. However, the consumer node in my
>> simulation does not receive the data packets if the producer is not
>> directly in the wireless range of the consumer node. The data are
>> considered unsolicited by the relay nodes. For your information, I have
>> modified the ndn-simple-wifi example to make this ad-hoc scenario. I have
>> changed the wifi mode to ad-hoc and increased the number of nodes to 9.
>> Is there anything I am missing? Is there any special NetDevice for
>> vehicular communication? For a single chunk of data, consumer generates one
>> interest that receives multiple time to the producer and the producer
>> respond multiple time accordingly but none of the data reaches the consumer.
>> My scenario
>> node 0                     consumer                node 2
>> node 3                     node 4                     node 5
>> node 6                     producer                  node 8
>> When I made node 0 as the producer, almost all the data packets
>> delivered. After, I make node 5 as the producer node, only 2 data packets
>> delivered against 20 Interest packets.
>> Thank you.
>> On Sat, Oct 31, 2020 at 8:31 PM DJEMAA Adel <a_djemaa at esi.dz> wrote:
>>> Hello,
>>> According to my small experience with wireless ad hoc scenario in
>>> ndnSIM, I think and I am sure that, all nodes send and receive packets
>>> (Interest and Data) on the same NetDevice Face (in my case 256), which is
>>> logically correct, since all nodes share the same wireless medium (for
>>> example LrWpan Or Wifi).
>>> For the PIT, it keeps the ID of the incoming face with the Interest
>>> prefix Name. In our case (ad hoc scenario), the ID of the incoming face is
>>> always the same, consequently, the sender of the interest packet couldn't
>>> be identified through the incoming Interest Face. However, when the the
>>> node sends back the Data Packet according to its PIT Entry (face 256), only
>>> the concerned node (s) among its neighbors (those who have asked for this
>>> Data packet i.e. PIT entry matches with prefix of the received Data packet)
>>> process this data packet, otherwise it is dropped (Unsolicited Data), and
>>> so on hop by hop until reaching the initial consumer.
>>> This is the basics of the wireless ad hoc scenarios in ndnSIM, as I have
>>> understood and confirmed by simulation in ndnSIM, and i Hope that will help
>>> you.
>>> Sincerely,
>>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Garanti
>>> sans virus. www.avast.com
>>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
>>> <#m_-1577891565979592351_m_6587483097763728456_m_6890541469200943740_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
>>> Le sam. 31 oct. 2020 à 09:22, Sonia via ndnSIM <ndnsim at lists.cs.ucla.edu>
>>> a écrit :
>>>> Adding to the above:
>>>> In wireless ad-hoc mode and broadcast strategy, every packet is
>>>> broadcasted on face 257 (same for all nodes) and receives the packet on
>>>> face 256 (same for all nodes). I am curious how the PIT works here? How can
>>>> the PIT keep the Interest traces and face number and how can it be used
>>>> while forwarding data back toward the consumer since there are only 2 faces
>>>> (one for sending Interest and one for receiving it)?  I will be very
>>>> grateful if I get some clue to solve my problem.
>>>> Thank you for your time.
>>>> On Fri, Oct 30, 2020 at 7:05 PM Sonia <zohanoor67 at gmail.com> wrote:
>>>>> Hi Everyone,
>>>>> I have a V2V scenario consists of 9 wireless ad-hoc nodes in a grid
>>>>> (no mobility currently) to check multihop communication. The consumer and
>>>>> producer node are both at the two extremes of the grid kind of topology. I
>>>>> and using the simple broadcast strategy similar to this example (
>>>>> https://github.com/4th-ndn-hackathon/ndnSIM-Mobile-Simulation-Package/blob/8de075fb57cf16358e6c408fac166c3c33f3664c/examples/ndn-v2v-simple.cpp ).
>>>>> All the Interest packets are delivered to the producer node via multipath
>>>>> communication and the producer nodes respond with data packets.
>>>>> However, the problem is that all the Data packets are deemed
>>>>> unsolicited and being dropped by the intermediate relay nodes.
>>>>> I also increased the Interest Life time from default 2 sec to 4 and 8
>>>>> sec to make it sure that the PIT entries are not expired due to congestion
>>>>> and long delay in the arrival of data packets. But it didn't work.
>>>>> Am I missing something?
>>>>> Your comments would be appreciated.
>>>>> Thank you
>>>>> Regards
>>>>> Sonia
>>>>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Virus-free.
>>>>> www.avast.com
>>>>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
>>>>> <#m_-1577891565979592351_m_6587483097763728456_m_6890541469200943740_m_-3064216562546481018_m_5295031873655921464_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
>>>> _______________________________________________
>>>> ndnSIM mailing list
>>>> ndnSIM at lists.cs.ucla.edu
>>>> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim
>>> --
>>> Mr DJAMA Adel
>>> Phd student
>>> Ecole Militaire Polytechnique
>>> Algiers 16046, ALGERIA
> --
> Mr DJAMA Adel
> Phd student
> Ecole Militaire Polytechnique
> Algiers 16046, ALGERIA
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndnsim/attachments/20201119/28507930/attachment.html>

More information about the ndnSIM mailing list