my instantaneous diagnosis would be that your PC is simply too slow at supplying samples. Incresing the rate of the strobe doesn't help -- the USRP sink is in burst mode and as such, will not complain if it's currently not sending samples. Underruns only happen if you're currently transmitting, but the samples come in too slow.
Now, the TX operations aren't all that CPU-intense; maybe your PC is simply a bit underpowered?
Best regards,
Marcus
PS: I took the freedom to de-tangle all the connections, so here's a picture of Abhinav's cleaned-up flow graph:
On 29.03.2016 17:21, Abhinav Jadon
wrote:
I was working on the transmitter of gr-ieee80211. I will first explain my setup and then go on to detail the problem.Hi ,
Setup :
X310 (transmitter) -> SMA cable -> Attenuator -> SMA cable -> X310 (receiver)
The transmitter runs the wifi_tx.grc. The receiver runs the wifi_rx.grc
The tx gain was intelligently set to 10 db . Same goes for rx gain.
The attenuation was set to -10db
The gains and attenuation value were set after conducting a small experiment where the transmitter was transmitting a sin wave of 0.8 amplitude and while the attenuation and tx gain were fixed, the value of rx gain was varied to see up till what value the received signal is not distorted.
The packet that was being transmitted by the transmitter was seen on a WX GUI scope sink and was consequently multiplied with an appropriate constant value to scale the max of the amplitude of the packet to 0.8.
The TX and RX chains were checked using the wifi_loopback flowgraph.
Now ,
The problem is that on the TX front, the flowgraph throws a lot of Underruns, my understanding is that the samples are not being created at the rate hardware is expecting them to be generated. The USRP block is configured on 20e6 sample rate in accordance with the Wifi Standard.
An educated guess is to increase the output of the flowgraph to the rate hardware is expecting. The message strobe block then needs to be configured to he expected rate but it exceeds the maximum pdu length.
I modified the tx and rx flowgraphs a bit, so here are the links to the flowgraphs :
Regards
Abhinav PS Jadon
2012122
Electronics and Communication Engineering Undergraduate
_______________________________________________ Discuss-gnuradio mailing list address@hidden https://lists.gnu.org/mailman/listinfo/discuss-gnuradio