How to raise the packet error rate in benchmark

GNU Radio
1

Hi everyone,

I’m doing a project on joint decode. But my problem is that in the lab
environment, the benchmark transmission is always good, there are
hardly any packet loss, so I can not show the advantage of joint decode.

I’ve tried to change the physical distance from transmission to receiver
attenna. Still works really well even put in opposite side of the
building.

I’ve tried to change the transmission power and transmission rate.
Inspired by some previous study, I learned that transmission power is
controlled by -tx-amplitude and -tx-gain, in which -tx-amplitude sets
the amplitude of the signal going into the DAC. RF gain is applied in
the daughtercard once the signal has been up converted. Now I’m using
0.2 amplitude and 35 gain. But I don’t understand, how come that the
smaller the power is, the better transmission performance there will be.
And also, why does power should be in the range of 0 to 1.0? How to
calculate transmission power using the -tx-amplitude and tx-gain?

Is there any other way that may cause more error in transmission beside
the ones I mentioned?

Thanks in advance! Any suggestion will be greatly appreciated!

Ada

2

Ada,

once the signal has been up converted. Now I’m using 0.2 amplitude and 35
gain. But I don’t understand, how come that the smaller the power is,
the
better transmission performance there will be. And also, why does power
should be in the range of 0 to 1.0? How to calculate transmission power
using the -tx-amplitude and tx-gain?
Better performance is not due to smaller power but it is due to the
non-linear effects of RF-amplifiers. As per my observations, Tx-Dsp
amplitude beyond 0.15 to 0.2 causes the USRP daughter-boards
Tx-amplification-stage to go into non-linear region thus error-rate of
PSK/QAM increases. Non-linear modulation e.g. FSK/GMSK are more robust
to
Amps non-linear effects. so u can experiment increasing TX-amplitude
both
for FSK-variants & PSK/QAM and observe performance.

-Adeel

3

Hi Adeel,

Thanks for your prompt reply!

I don’t quite understand your last email. If I want to test the
performance of transmission under different power allocation. What
should I do? Can you please be more specific about this?

Currently, what I’m doing is keep the tx-amplitude 0.2, and change the
gain from 35 to 65. You mean in this range, the power is not liner to
gain change, right? If so, is there any way to control or influence the
power change?

Thank you!

Regards,
Ada

Date: Wed, 24 Jul 2013 06:12:15 +0500
Subject: Re: [Discuss-gnuradio] how to raise the packet error rate in
benchmark
From: [email protected]
To: removed_em[email protected]; [email protected]

Ada,

once the signal has been up converted. Now I’m using 0.2 amplitude and
35 gain. But I don’t understand, how come that the smaller the power is,
the better transmission performance there will be. And also, why does
power should be in the range of 0 to 1.0? How to calculate transmission
power using the -tx-amplitude and tx-gain?
Better performance is not due to smaller power but it is due to the
non-linear effects of RF-amplifiers. As per my observations, Tx-Dsp
amplitude beyond 0.15 to 0.2 causes the USRP daughter-boards
Tx-amplification-stage to go into non-linear region thus error-rate of
PSK/QAM increases. Non-linear modulation e.g. FSK/GMSK are more robust
to Amps non-linear effects. so u can experiment increasing TX-amplitude
both for FSK-variants & PSK/QAM and observe performance.

-Adeel

On Tue, Jul 23, 2013 at 9:25 PM, yeran [email protected] wrote:

Hi everyone,

I’m doing a project on joint decode. But my problem is that in the lab
environment, the benchmark transmission is always good, there are
hardly any packet loss, so I can not show the advantage of joint decode.

I’ve tried to change the physical distance from transmission to receiver
attenna. Still works really well even put in opposite side of the
building.

I’ve tried to change the transmission power and transmission rate.
Inspired by some previous study, I learned that transmission power is
controlled by -tx-amplitude and -tx-gain, in which -tx-amplitude sets
the amplitude of the signal going into the DAC. RF gain is applied in
the daughtercard once the signal has been up converted. Now I’m using
0.2 amplitude and 35 gain. But I don’t understand, how come that the
smaller the power is, the better transmission performance there will be.
And also, why does power should be in the range of 0 to 1.0? How to
calculate transmission power using the -tx-amplitude and tx-gain?

Is there any other way that may cause more error in transmission beside
the ones I mentioned?

Thanks in advance! Any suggestion will be greatly appreciated!

Ada

Discuss-gnuradio mailing list

[email protected]

https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

4

Currently, what I’m doing is keep the tx-amplitude 0.2, and change the
gain from 35 to 65. You mean in this range, the power is not liner to
gain
change, right? If so, is there any way to control or influence the power
change?
I was mentioning about the that total Tx-Power. U can control Tx-power
by
varing the value of tx-dsp-amplitude from 0-1 in daughter-boards that
don’t
have variable tx-gain e.g. RFX2400 otherwise its the combination of both
tx-dsp-amplitude and tx-gain.

-Adeel

5

In ur GR flowgraph, just add a const-multiply block before USRP-Sink and
multiply the Tx signal with it. Set its value between 0-1.

-Adeel

6

On Sat, Jul 27, 2013 at 12:03 AM, Adeel A. [email protected]
wrote:

I’m using RFX 2400 for experiment, there are no parameter
benchmark
tx-dsp-amplitude and tx-gain.

Tx-amplification-stage to go into non-linear region thus error-rate of

amplitude of the signal going into the DAC. RF gain is applied in the

Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page

Changing this amplitude is a way to change the transmit power, but
that’s not really the purpose of this amplitude. The tx-gain is the
better way to change the tx power. At one point you mentioned you were
using 35dB gain. Why don’t you lower that to 0 and then increase the
gain until you have sufficient reception?