Friis Propagation loss

[peyman]

Hi all

I have 4 nodes and I am using Friis Propagation loss model between
them. one question is when I use this model, one node can receives all
the packets until one distance and after that distance receiver
doesn't receive anything. I want to have for example 80 % loss but
when I am changing the distance, the loss probability is 0% or 100%.
it means that receiver receives all the packer or nothing !!
another question is is there any way to change parameters of loss
model for each link??

Best regards

[Konstantinos]

Dear Peyman,

Friis propagation model is a deterministic one. Meaning that pathloss is only affected by the distance.
In your scenario, the only parameters that affect the reception of the packet are the pathloss and the noise (constant). This means that the SINR value of the received packet is only affected by the distance that the two nodes have.

You have to add some "stochastic" parameter (randomness) in order to have tis 80% you mention. If you use another model (e.g. Nakagami) that is not deterministic, you can have such behavior.
 
Regards,
Konstantinos

[Sascha Alexander Jopen]

Hi,

with Friis propagation loss model, this is not possible. This model
simply calculates the signal power at the receiver, depending on the
distance, frequency and antenna gains. This is deterministic and as
long as you do not change the frequency or antenna gains, the received
signal strength decreases with the distance only, eventually dropping
below your receive threshold. So your nodes are either in
communication range or not.
What you want is another or an additional fading model, which is
probabilistic, modeling multipath propagation, interference and the
like. The only such model implemented to my knowledge in ns3 is the
Nakagami fading model. With special parameters, this model equals the
Rayleigh fading model, see the model documentation for details.
However, which model you use should strongly depend on the scenario
you are modeling. Outdoor scenarios typically use different models
than indoor scenarios.
You should keep this in mind, when using the Friis free space
propagation loss model as well.

Regards,
Sascha

[Peyman Pahlavani]

Thank you for your response,

I tried by Nakagami fading model but again I have the same problem, I also changed the m parameter but I could not solve the problem.

Best Regards
Peyman

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To post to this group, send email to ns-3-...@googlegroups.com.
To unsubscribe from this group, send email to ns-3-users+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/ns-3-users?hl=en.

[Peyman Pahlavani]

sorry some packets are lost but in deterministic manner. for example 5 packets are always lost when a node send 10 packets.

Best regards

[Konstantinos]

In order to get accurate results, you have to increase the number of iterations (e.g. send 1000 packets and get average). If you are only sending 10, then you can't get proper statistical results.

[Peyman Pahlavani]

Hi,

Thank you for your response.
 I sent 10000 of packets and I did the simulation several times.
 but the receiver receives 5819 packets in each iteration,
it seems that there is no randomness in Nakagami fading model!!

Best regards

On Tue, Feb 21, 2012 at 6:41 PM, Konstantinos <dinos.k...@gmail.com> wrote:

In order to get accurate results, you have to increase the number of iterations (e.g. send 1000 packets and get average). If you are only sending 10, then you can't get proper statistical results.

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.

To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/mUTozt2NTvUJ.

[Luca Costantino]

Il 22 febbraio 2012 09:55, Peyman Pahlavani <peyma...@gmail.com> ha scritto:
> it seems that there is no randomness in Nakagami fading model!!

did you try changing the seed number?

--
Chiave pubblica http://luca.costantino.googlepages.com/luca.costantino.asc

Prima di tutto vennero a prendere gli zingari e fui contento, perché
rubacchiavano.
Poi vennero a prendere gli ebrei e stetti zitto, perché mi stavano antipatici.
Poi vennero a prendere gli omosessuali, e fui sollevato, perché mi
erano fastidiosi.
Poi vennero a prendere i comunisti, e io non dissi niente, perché non
ero comunista.
Un giorno vennero a prendere me, e non c’era rimasto nessuno a protestare.
(Martin Niemöller)

[Peyman Pahlavani]

No I didn't How can I do that ??

2012/2/22 Luca Costantino <luca.co...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.

[Luca Costantino]

Il 22 febbraio 2012 10:13, Peyman Pahlavani <peyma...@gmail.com> ha scritto:
> No I didn't How can I do that ??

http://www.nsnam.org/docs/release/3.13/doxygen/classns3_1_1_seed_manager.html

[Peyman Pahlavani]

Thank you very much.
 I set the different seed for each iteration and I got different result.