RSSI estimation

GNU Radio
1

Hello,
Anyone know if it is correct to roughly estimate RSSI information
from I/Q values. For example is it correct to say RSSI of a message is
equal
to 10log10(I^2-Q^2). Otherwise, what is the correct way to obtain RSSI
value
in gnuradio?

Thanks,
Bin
Winlab

2

Hello Bin,

My group considered doing this about 4-5 months ago. Things may have
changed now, but at that time it was not possible (or at least not
straight
forward to me) to get the gain from the AGC.

From my understanding, the AGC (automatic gain control) makes getting
RSSI
readings difficult because your signal is getting amplified before it
gets
sampled. So if you don’t know the gain the AGC sets (which varies with
time) it’s hard to do RSSI.

Devin

3

On 04/27/2010 09:04 AM, bin zan wrote:

Hello,
Anyone know if it is correct to roughly estimate RSSI
information from I/Q values. For example is it correct to say RSSI of a
message is equal to 10log10(I^2-Q^2). Otherwise, what is the correct way
to obtain RSSI value in gnuradio?

You would want to average over many samples. You will also need to take
account of the current gain setting.

Matt

4

Thanks Devin and Matt.
From your reply can I say if I don’t use AGC and with a fixed value of
gain,
then I can calculate the RSSI value.

Bin Zan

5

On 04/27/2010 06:07 PM, bin zan wrote:

Thanks Devin and Matt.
From your reply can I say if I don’t use AGC and with a fixed value of
gain, then I can calculate the RSSI value.

Yes