Method to improve the detection sensitivity of USRP1 with LFRX

GNU Radio
1

Dear all,

I’m using LFRX to detect a -120dBm signal. According to my measurement,
the lowest power of signal that LFRX is able to detect is -90dBm. I
tried deploy a preamplifier to amplify the received signal by certain dB
in order to reach the -90dBm, but this turned out that noise is
amplified as well and the signal still merges in noise, undetectable. I
also tried to set proper bandwidth for the lowpass filter in order to
eliminate noise and avoid the signal distortion to the largest extend,
but this still cannot allow this weak signal be detected.

The flow-graph of my receiver is:

usrp_source_c --> Blackman lowpass filter -->
file_sink(gr.sizeof_gr_complex)

The signal that I wish to detect is a 13-bit Barker code with 40 us chip
rate. The samples frequency at receiver side is 250kHz which gives the
decimation rate 256. Regarding with the Blackman lowpass filter, the
sampling frequency for this filter is 35kHz, and the cutoff frequency is
17kHz, transition_width is 0.1kHz.

Really appreciate any of your help!

Yan

2

On 03/28/2011 06:41 PM, Yan N. wrote:

The signal that I wish to detect is a 13-bit Barker code with 40 us chip rate.
The samples frequency at receiver side is 250kHz which gives the decimation rate

  1. Regarding with the Blackman lowpass filter, the sampling frequency for this
    filter is 35kHz, and the cutoff frequency is 17kHz, transition_width is 0.1kHz.

Really appreciate any of your help!

What’s the carrier frequency?

At -120dBm, you do need a low-noise amplifier with sufficient gain to
overcome the noise of the ADC.
Higher-noise amplifiers might indeed have the effect you’re seeing.
I’ve used the ERA-3 series from
Mini-Circuits sucessfully for low-noise HF reception at around 38MHz
on a BASIC_RX.

–
Principal Investigator
Shirleys Bay Radio Astronomy Consortium

3

Really appreciate your reply. The carrier frequency that I use is 6MHz.
I would like to try the preamplifier that you recommended.
I got another problem which may clear the reason why I met this problem.
I used a signal generator to produce a 6.01MHz sinusoid wave. After down
conversion by 6MHz, it’s supposed to show the 10kHz in the spectrum.
However, the spectrum that I got gives a really big 0 frequency
component beside the 10kHz component. It seems that a DC offset occurred
during signal reception. Is it right? If so, what causes this problem?
How to eliminate the DC offset?

Many appreciation!

Yan