Wifi interference

[Michaël]

Hello,

I'm trying to implement a scenario where I have multiple STA nodes, each connected to an AP node. How exactly do I need to program this to have interference between the different wifi connections? 

Let's say staNode0 is connected to apNode0 through wifi, and staNode1 is connected to apNode1 through wifi. Originally I implemented these 2 wifi set-ups apart and tried changing parameters (like same channel number in the WifiPhy for example), but when i moved the 2 apNodes closer to each other, the throughput remained exactly the same (no interference). I also tried using the same Ssid or the same wifi ap configuration on the 2 apNodes and same wifi sta configuration on the 2 staNodes, but i had the same result.

Any help would be greatly appreciated.

[Konstantinos]

Hi,

There is an example in /examples/wireless/wifi-simple-interference.cc demonstrating interference.

Note that inter-channel interference is not implemented in current wifi module (yanswifiphy). There are discussions about this issue. 

[Michaël]

Hi,

Thanks for the fast reply. I already looked at the examples, but they don't really seem to help me any further since my scenario is different. I also read that cross-channel interference is not yet implemented, but that is why I also mentioned I tried using the same wifi configuration or ssids for the different ap nodes. Am I not correct that interference between 2 ap nodes for the same ssid should be possible?

Op zondag 30 november 2014 23:06:38 UTC+1 schreef Konstantinos:

[Konstantinos]

Hi,

You are correct that there should be interference. Even with different ssid if you are working on the same channel.

However, it is not only the channel number in the PHY, it has to be the same PHY/Channel instance (you are working with objects).

So, if you have multiple "wifiChannel.Create()" statements, that creates new channel instances, hence no interference.

[Michaël]

Hi,

Thanks once again for your help on this matter. Since I couldn't really find an example where the scenario I needed was implemented, and I'm also quite new to NS-3, I tried some possibilities myself, like using the same channel number, same ssid, and even the same Phy, but I did indeed mistakenly use the wifiChannel.Create() statement for each of the connections. I will modify my code tomorrow, and hopefully my problem will be solved. 

Op zondag 30 november 2014 23:37:17 UTC+1 schreef Konstantinos:

[Michaël]

Hi,

I have a question related to this problem. I'm trying to investigate the performance of some different configurations, and without cross-channel interference this will of course be limited. Browsing through some of the topics in this group, I noticed that using the Spectrum module would be able to provide a solution, but since this requires a significant amount of programming, that isn't really an option for me.  

Another (coarse but easy) solution for this problem that was mentioned in this group was modifying the YansWifiChannel::Send method by adding some attenuation to the receiver power (so I'm assuming adding some attenuation to the double rxPowerDbm variable). By looking at the following paper:

Eduard Garcia Villegas, Elena Lopez-Aguilera, Rafael Vidal, Josep 
Paradells, "Effect of adjacent-channel interference in IEEE 802. 1 
WLANs" 

and more specifically table I, this looks like a good solution. I know it would be a coarse solution, but if I were to use the data from table I, would the results I get be somewhat realistic?

Another related question: If i want to compare the throughput of configurations with wifi connections on the same channel to configurations with wifi connections on different channels, is it sufficient to only set a different channel number for these different connections (and still use the same phy and channel instance), or should I also create different phy and channel instances?

Thanks in advance

[Scapegoat]

Hi, Michael. I also have a problem with this issue, but my scenery which depend on multi-interface multi-channel WLANs. So, i hope you can share your solution with me. Thank you very much！

                       Best Regards
在 2014年12月3日星期三UTC+8下午9时46分19秒，Michaël写道：