Inconsistency in the PER versus distance using USRP

Hello, everybody!

I am using two USRP to find the average PER at a certain distance, but
the average PER at a certain distance seems to be varying every day.
Yesterday, the PER was around 5% when two usprs were 10 feet apart, but
today the PER was around 2.5%Â which has been diminished by 50%! How
does this come? I am really shocked? My wireless environment is
described below:

In door environment; the temperature are the same in these two
experiments; center frequency is 2479MHz because on one hand, I want to
avoid the 2400MHz to 2483MHz band which is the working band of 802.11,
on the other hand the bandpass filter on the 2400MHz daugterboard (2400M
to 2483M) and the bandwidth of USB port which is 8MHz together forced me
to chose the center frequency to be at most 2479MHz; all other
conditions in these two experiments are the same: chairs and desks are
fastened, not being moved, etc.

I am very confused about the different results USRP gave me in these two
days. Could some one tell me the reason? Any hint will be highly
appreciated!!!

Thanks a lot!

Bill

Bill S. wrote:

In door environment; the temperature are the same in these two
experiments; center frequency is 2479MHz because on one hand, I want to
avoid the 2400MHz to 2483MHz band which is the working band of 802.11,
on the other hand the bandpass filter on the 2400MHz daugterboard (2400M
to 2483M) and the bandwidth of USB port which is 8MHz together forced me
to chose the center frequency to be at most 2479MHz; all other
conditions in these two experiments are the same: chairs and desks are
fastened, not being moved, etc.

Didn’t you just answer your own question? You’re operating within
802.11 interference range. The frequency band of channel 12 is
2456-2478MHz. You’re sitting at 2479MHz and with 8MHz bandwidth, your
bottom 4MHz are within 802.11 interference range. Additionally, you
could use usrp_fft.py to actually inspect the channel and determine if
there is any noise on it.

  • George

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Nov 20, 2008, at 6:40 PM, George N. wrote:

most 2479MHz; all other conditions in these two experiments are the
same: chairs and desks are fastened, not being moved, etc.

Didn’t you just answer your own question? You’re operating within
802.11 interference range. The frequency band of channel 12 is
2456-2478MHz. You’re sitting at 2479MHz and with 8MHz bandwidth,
your bottom 4MHz are within 802.11 interference range.
Additionally, you could use usrp_fft.py to actually inspect the
channel and determine if there is any noise on it.

To be fair, assuming 20 MHz wifi, he should be fairly interference
free. Channel 11 (2.462G) is pretty well attenuated at 2.479G
according to the standard. I suspect you’re going to have to dig in
Bill. Note that multipath can have a huge effect, as well as frequency
offsets and other variables, like the output power changing on the
card. There’s usually not an obvious easy answer for these things
without digging into the RF.

If I move my leg I can knock the PRR from 100% to 5% easily – using
commercial WiFi.

  • -Dan
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.8 (Darwin)

iEYEARECAAYFAkkmIZQACgkQy9GYuuMoUJ6UJwCfV73j1smggESY+WKqn6C8UzhH
K1EAnAgnSmZ0YT2hrW3tmZos2MPONkyY
=Vigv
-----END PGP SIGNATURE-----


From: Dan H. [email protected]
To: George N. [email protected]
Cc: Bill S. [email protected];
[email protected]
Sent: Thursday, November 20, 2008 9:48:52 PM
Subject: Re: [Discuss-gnuradio] inconsistency in the PER versus distance
using USRP

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Nov 20, 2008, at 6:40 PM, George N. wrote:

Bill S. wrote:

In door environment; the temperature are the same in these two experiments; center frequency is 2479MHz because on one hand, I want to avoid the 2400MHz to 2483MHz band which is the working band of 802.11, on the other hand the bandpass filter on the 2400MHz daugterboard (2400M to 2483M) and the bandwidth of USB port which is 8MHz together forced me to chose the center frequency to be at most 2479MHz; all other conditions in these two experiments are the same: chairs and desks are fastened, not being moved, etc.

Didn’t you just answer your own question? You’re operating within 802.11 interference range. The frequency band of channel 12 is 2456-2478MHz. You’re sitting at 2479MHz and with 8MHz bandwidth, your bottom 4MHz are within 802.11 interference range. Additionally, you could use usrp_fft.py to actually inspect the channel and determine if there is any noise on it.

To be fair, assuming 20 MHz wifi, he should be fairly interference free.
Channel 11 (2.462G) is pretty well attenuated at 2.479G according to the
standard. I suspect you’re going to have to dig in Bill. Note that
multipath can have a huge effect, as well as frequency offsets and other
variables, like the output power changing on the card. There’s usually
not an obvious easy answer for these things without digging into the RF.

If I move my leg I can knock the PRR from 100% to 5% easily – using
commercial WiFi.

  • -Dan
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.8 (Darwin)

iEYEARECAAYFAkkmIZQACgkQy9GYuuMoUJ6UJwCfV73j1smggESY+WKqn6C8UzhH
K1EAnAgnSmZ0YT2hrW3tmZos2MPONkyY
=Vigv
-----END PGP SIGNATURE-----

Thank you for your help, you gave me a better understanding about the
wireless environment!

Bill