802.11 Application Framework Ouput

[thewifiguy]

Hey everybody,

I recently started to measure the Transmit Power of the 802.11 Application Framework with a USRP RIO 2953R.

My measurement setup was as follows (It is depicted in the attachment):

802.11 Application Framework  v1.1 + USRP RIO 2953R - Cable + 30dB Attenuator - Rohde&Schwarz FSV Analyzer in Channel Power Measurement Mode

The 802.11 Application Framework ran the 802.11ac 20 MHz Channel with a center frequency of 2.5 GHz

In the GUI of the Application Framework the TX power can be changed manually and it is given in dBm. According to the Device Specifications of the USRP RIO 2953, the transmit power goes up to 20dBm.

Two things stand out:

1. The values that can be entered manually do not correlate to expected dBm value. There is an offset of about -15dBm.

2.  The maximum measured TX Power is 9 dBm.

I attached my measurement results.

My question is:

How can I reach the maximum output power of 20 dBm? Or did I overlook something?

Thanks in advance.

Best,

thewifiguy

[komisk86]

Hi thewifiguy.

My name is Kiril from AE israel team and I will try to assist you with your issue.

first of all I have reproduced your issue with USRP 2940R and VST ( instead of SA ) and 30 dB attenuator .

you are right , there is a gain offset between the manual value and the real measured value. this offset is depended on your central frequency .

in lower central frequencies you will have a smaller offset .

this offset occured due the fact that the USRP is uncalibrated device and this is an

official answer.

the USRP 2950R is capble to get the 20dBm value of power, and this should be possible programmatically as a workararound.

in generally my suggestion is to manipulate the modulation data amplitude

in order to increase the transmitted power .

in 802.11 design it is harder because the modulation data is on the FPGA side  .

the file is called Data Modulator .gcdl and you can change it basically.

but this is my proposal of workaround .

what I did , is changed the TX Async example in order to get higher power level.

I changed the amplitude of the waveform GVI in the Host side and I have received -15dBm channel power which is the actaual 15dBm of real power. 

it should be the same or similar with the application framework .

please let me know if you have additional questions.

Best regards.

Kiril.

[thewifiguy]

Hi Kiril,

thanks for your quick response and sorry for my late answer.

Your answer, that the USRP is an uncalibrated device explains some behaviour of my measurements.

Unfortunately, I couldn't transfer your TXasync example to the Application Framework.

Originally, I was searching for a way to calculate a SNR. Is there any way to do this with an uncalibrated device like the USRP or am I completely lost here?

Best,

Simon

[Hey.itsDale]

I think you will have an easier time measuring SNR than absolute power since it is a ratio of powers. The measurement will be accurate to the extent that the offset in power measurement is constant across frequency.

The simplest way to measure SNR would probably be to observe the power spectrum of the signal and determine it manually (see section 4 here):

http://www.ni.com/white-paper/3359/en/#toc4

If you are only interested in the SNR within the passband of your 802.11 signal, things become more complicated. In general one way to do this is to project the signal you sent on to the signal you recieved (correlate them). It will help to do this over a range of time offsets and search for a maximum (there will be some uncertainty in propogation delay). Then subtract this component from the sigal. The ratio of the power of the extracted component to the power of the remaining signal is SNR.