Monitor behaviour of a neighbouring node

[faisal Lone]

Hi all. I am new to NS-3 and need some help regarding monitoring behaviour of a neighbouring node in a VANET. I want each of the nodes to measure the packeks received, dropped and forwarded of each of their neighbours. How can i configure a node to do that?

I also want to add RSUs to the VANET scenario. IS creating static nodes the way to do it? and how to place these RSUs at specific location in SUMO using Openstreet maps.

Any help would be greatly appreciated.

Regards

Faisal

[Tommaso Pecorella]

For the second question I have no clear answer - placing them statically in a given point looks logical, but I know nothing about SUMO.

About the first question... you don't. Or better: you could, but you should not do it.

Now, I know that some papers claim that measuring the number of packets received, dropped, and forwarded by a neighbour is an excellent way to do stuff, but the problem is... you #@&%$ can't. It's simply make-believe.

Let's analyse the problem:

• Forwarded (or sent): simply listen to the channel and, if you're close enough, you can receive what your neighbour is sending. Caveat: you can't receive if you're transmitting, and you could loose some packets due to channel errors. Plus, you'll use.a LOT of energy.
  
• Dropped: you can't measure that. At best you could measure the number of received and not forwarded. IF you could measure the received packets - and you can't, see below.
• Received: YOU CAN NOT MEASURE IT. Unless the two nodes are so close to be practically in the same place, and even in this case you can't.

Now, mind, I'm not talking just about simulations. This is the harsh reality of the physical world.

When a node send a packet, a receiver will receive a signal, modified by the channel. The ability of the receiver to decode the signal depends on how good are the receiver antennas, the circuits, the decoding algorithm, and so on. Even without too many details, the point is: if you have a transmitter and two receivers, the two receivers will have a different channel (they'll receive different signals).

This can (and usually does) mean that one will be able to decode the packet, and the other no. The most known effect is called "hidden node problem", and it is far more common than one might expect.

Hence, can you measure what a neighbour is receiving? Absolutely no. You'll always either overestimate or underestimate the numbers depending on the respective positions of the 3 nodes (Tx, Rx1, Rx2).

As a consequence, every time you see a paper, or a simulation, based on these numbers... you're reading a make-believe (or a paper-ware).

Sorry if I did just crush a good chunk of existing literature, but this is not my opinion, it's physics.

PS: can you devise a way to have a node measure these quantities by itself (I mean, for itself, not for a neighbour), and then broadcast these?

Yes, but the node could lie.

[faisal Lone]

Thanks a lot for the prompt reply....any suggestion on how to measure behaviour of nodes in a network other than being monitored by a neighbouring node? Thanks again for the help

[faisal Lone]

Can we have multiple NICs on a node where one is used for reception and transmission whereas the other is used for sniffing?

[Tommaso Pecorella]

You can have as many NICs you want on a node (both real and ns-3), but then again... a second NIC will be able to monitor the other NIC on the same node, not the ones of neighbours.

About monitoring neighbours, you can do it, but you'll have to take into account the fact that you're not listening to the "correctly". You might apply some statistical models to correct the data, but they'll be inevitably have to make assumptions on how many nodes the monitoring node isn't able to detect.

Now I'll go technical... sorry.

If a monitoring node wants to monitor a neighbour, it should have an antenna with a sensitivity so high to be able to see (reliably) any node that is 2 hops away from it. That, of course, assuming that there are no corners or objects in between the 2-hops away nodes.

So, let's assume that the environment does not contain corners, or corridors, or other situations where the radio propagation is heavily impaired. Kinda unrealistic, but let's assume that.

Let's also assume that the monitor node can detect the direction of the signal, and, through some magic, it can guess what's the signal received by the neighbour. This involves also estimating the channel matrix at the neighbour, and to do that you'll also have to know the antenna gain and type of the neighbour. 

If you do this assumption, and the monitor node is FAR more powerful than a normal node, it could monitor a neighbour decently (not perfectly, but decently). You'll not have the exact number of packets received by the neighbour, but you'll have a decent guess. 

Are these assumptions realistic? Absolutely not, especially if you think that the neighbour is malicious.

So... you see that in a practical scenario if the monitor node receive a packet you don't know if the neighbour will receive it, and if the monitor node does not receive a packet you can't tell if the neighbour did receive a packet.

Hence, you can't monitor the neighbour.

So, is it monitoring neighbours a completely idiotic idea?

Well, no. But you'll have to do it in way more complex ways. Here's where research comes into play: monitor the neighbour's behaviour taking into account that the only thing you really know is how many packets they did forward (packets sent but that are not originated by the node, i.e., with a mismatch between the sender MAC address and the sender IP address).

How? Well, I don't know - otherwise I'd be writing a paper right now.

[faisal Lone]

Thanks a ton for the reply..I am getting what your saying..maybe I ll change my approach. Thanks a ton again..have some some queries...will post them too.

--
Posting to this group should follow these guidelines https://www.nsnam.org/wiki/Ns-3-users-guidelines-for-posting
---
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ns-3-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ns-3-users/1fb71752-7753-478e-84de-005c5fff53f0n%40googlegroups.com.