# Re: GSM spectrum: Invisible 200kHz carriers?

Operator,

Are you saying that you are seeing 4MHz wide signals? Or are you seeing
~200KHz wide signals seperated by 4MHz?

If you are seeing 4MHz wide signals, they are not GSM. There is most
likely
another technology being transmitted.

If you are seeing ~200KHz wide signals spaced by 4MHz, these most likely
are
GSM signals. The spec states that the spacing between adjacent channels
is
200KHz, but that does not mean that an operator will place all of their
channels next to each other. In fact, it would be bad if they did this.
Keep in mind that the spacing between channels is 200KHz, but the
occupied
bandwidth of a GSM signal is ~270KHz. This means that if GSM channels
are
placed next to each other, ICI will increase significantly. Typically
what
is seen is a very sparse scattering of channels across the band.

Hope this helps,
Ben
On Jun 2, 2011 4:49 PM, “TheOperator” [email protected] wrote:

Ben W. wrote:

Are you saying that you are seeing 4MHz wide signals? Or are you seeing
~200KHz wide signals seperated by 4MHz?I see about 200kHz wide signals
spaced 4MHz.

Ben W. wrote:

is seen is a very sparse scattering of channels across the band.Thanks for
the explanation, this makes sense. The channels are not exactly 200kHz
wide however (see screenshots), but according to the frequencies, they
have to be GSM.

I used this little system along with some GUI elements:
http://old.nabble.com/file/p31763653/gsm-900-blocks.png

The resulting spectrum looks like this:
http://old.nabble.com/file/p31763653/gsm-900.png

## Such a peak has the following shape. Most of the power is distributed within ~50kHz. Do these smaller peaks have a special meaning?http://old.nabble.com/file/p31763653/gsm-900-zoomed.png

View this message in context:
http://old.nabble.com/GSM-spectrum%3A-Invisible-200kHz-carriers--tp31761278p31763653.html
Sent from the GnuRadio mailing list archive at Nabble.com.

Alexander C. wrote:

Hi,

This doesn’t look like a GSM signal. There some pictures of GSM spectrum
captured with USRP on the internet. I don’t have a link near my hands, but
you can probably google them easily.

You are probably right. I found signals that look better:
http://old.nabble.com/file/p31807451/949M.png

## Do you know how the other peaks (in my earlier post) can be interpreted? They appear quite regularly in intervals of 4 MHz…

View this message in context:
http://old.nabble.com/GSM-spectrum%3A-Invisible-200kHz-carriers--tp31761278p31807451.html
Sent from the GnuRadio mailing list archive at Nabble.com.

Hi,

This doesn’t look like a GSM signal. There some pictures of GSM spectrum
captured with USRP on the internet. I don’t have a link near my hands,
but
you can probably google them easily.

Regards,
Alexander C…
On Jun 3, 2011 4:21 AM, “TheOperator” [email protected] wrote:

are
GSM signals. The spec states that the spacing between adjacent channels
is
200KHz, but that does not mean that an operator will place all of their
channels next to each other. In fact, it would be bad if they did this.
Keep in mind that the spacing between channels is 200KHz, but the
occupied
bandwidth of a GSM signal is ~270KHz. This means that if GSM channels are
placed next to each other, ICI will increase significantly. Typically
what
is seen is a very sparse scattering of channels across the band.Thanks
for
Such a peak has the following shape. Most of the power is distributed
within
~50kHz. Do these smaller peaks have a special meaning?
http://old.nabble.com/file/p31763653/gsm-900-zoomed.png

View this message in context:
http://old.nabble.com/GSM-spectrum%3A-Invisible-200kHz-carriers--tp31761278p31763653.html