The statistical model in the MAC is based on estimating the number of nodes transmitting and the bandwidth utilization. That info is used to determine collisions (direct neighbors, hidden neighbors, tx busy, etc.).
For unicast, the default retry limit is set to 2 which is probably why you are seeing little to no loss. You can reduce that number to 1 or 0 or send multicast traffic to see if you start to see packets being discarded. There are statistics you have access through the emanesh via the control port (see the documentation) that will show you packets discarded and why. Steve's response showed the table for broadcast packets and there is also one for unicast packets. There are also much more stats you have access to which may or may not be of interest to you. kb Kaushik B. Patel Adjacent Link LLC CEO, Director of Engineering On 05/08/2014 01:23 PM, Anders Nilsson Plymoth wrote: > Hi Steven, > > Thank you for the explanation, it helped. > > What I meant with inband interference, was interference from other 802.11 > nodes on the same frequency (same SSID, BSSID). Like you said, setting > noisemode to none, does increase contention and channel utilization which > is actually what I wanted. I was able to observe this using a TCP iperf > stream, but I was actually expecting some packets to be dropped due to the > interfering nodes saturating the channel. I don't observe any unicast > packet drops at all, which seems a bit unrealistic as it indicates that the > MAC (CSMA/CA) performs perfectly. I will degrade the channel and introduce > some hidden node scenarios and try again. > > One more question. Are there any statistics available that shows 802.11 > retransmissions? Or lost ACks? > > Thanks, > Anders > > > On Wed, May 7, 2014 at 8:48 PM, Steven Galgano > <[email protected]>wrote: > >> Anders, >> >> When using the 802.11 radio model, the physical layer 'noisemode' >> parameter can >> only be set to 'none' or 'outofband'. The 802.11 radio model uses a channel >> utilization and neighbor estimator statistical model for in-band >> interference. >> >> If you checked your logs you should have seen the following message when >> 'noisemode' was set to 'all' for each received packet: >> >> 23:06:47.679146 ERROR MACI 001 IEEEMACLayer upstream EOR processing: >> spectrum service reporting signal in noise. This is an unallowable noise >> mode. Valid PHY noise modes are: none and outofband. >> >> You can also check the BroadcastPacketDropTable0 and >> UnicastPacketDropTable0 to >> see why your packets are not completing. In this case, the 'Bad Control' >> column will >> increment for each received packet indicating there is an internal control >> message issue - the physical layer is configured incorrectly. >> >> emanesh (node-1:47000)] ## get table 1 mac BroadcastPacketDropTable0 >> nem 1 mac BroadcastPacketDropTable0 >> | NEM | SINR | Reg Id | Dst MAC | Queue Overflow | Bad Control | Bad >> Spectrum Query | Flow Control | Duplicate | Rx During Tx | Hidden Busy | >> | 2 | 0 | 0 | 0 | 0 | 130 | 0 >> | 0 | 0 | 0 | 0 | >> | 3 | 0 | 0 | 0 | 0 | 129 | 0 >> | 0 | 0 | 0 | 0 | >> | 4 | 0 | 0 | 0 | 0 | 129 | 0 >> | 0 | 0 | 0 | 0 | >> | 5 | 0 | 0 | 0 | 0 | 131 | 0 >> | 0 | 0 | 0 | 0 | >> | 6 | 0 | 0 | 0 | 0 | 130 | 0 >> | 0 | 0 | 0 | 0 | >> | 7 | 0 | 0 | 0 | 0 | 68 | 0 >> | 0 | 0 | 0 | 0 | >> | 8 | 0 | 0 | 0 | 0 | 71 | 0 >> | 0 | 0 | 0 | 0 | >> | 9 | 0 | 0 | 0 | 0 | 71 | 0 >> | 0 | 0 | 0 | 0 | >> | 10 | 0 | 0 | 0 | 0 | 73 | 0 >> | 0 | 0 | 0 | 0 | >> >> If you want to use an 802.11 network to jam another 802.11 network and both >> networks are set to the same center frequency you will need to set >> different >> physical layer subids. Otherwise you only have one network from a MAC >> perspective and are just increasing the channel utilization and contention >> of >> that network. >> >> https://github.com/adjacentlink/emane/wiki/Physical-Layer-Model#subid >> >> -- >> Steven Galgano >> Adjacent Link LLC >> www.adjacentlink.com >> >> >> On 05/07/2014 07:30 PM, Anders Nilsson Plymoth wrote: >>> Hi, >>> >>> How do I emulate in band interference using the IEEE 802.11abg model? >>> >>> If I have RF Pipe inteference nodes on the same frequency etc using >>> outofband noisemode I get the expected results and lowered SINR. >>> >>> However, when I switch to ieee802abg for the interference nodes, and >> using >>> noisemode *all*, I receive *nothing*. If I change to noisemode *none, *I >>> receive everything. This is very unrealistic, and if I understand the PHY >>> documention correctly I should be able to use in band interference. >>> >>> Thanks, >>> Anders >>> >>> >>> >>> _______________________________________________ >>> emane-users mailing list >>> [email protected] >>> http://pf.itd.nrl.navy.mil/mailman/listinfo/emane-users >>> >> > > > > _______________________________________________ > emane-users mailing list > [email protected] > http://pf.itd.nrl.navy.mil/mailman/listinfo/emane-users > _______________________________________________ emane-users mailing list [email protected] http://pf.itd.nrl.navy.mil/mailman/listinfo/emane-users
