On Thu, May 01, 2008 at 11:54:15AM +0000, steve wrote: > Hi, > > We are trying to understand how to implement GSM channel hopping using > the USRP1. > > Reading the docs and specs I think there might be a way of doing it without > any fpga modification. Comments are welcome if there is a mistake in my > thinking. > > Inband-Signaling gives us very precise timestamps on received packets: > > www.gnuradio.org/trac/browser/gnuradio/trunk/usrp/doc/inband-signaling-usb > > We can also use it to have certain commands executed on the fpga at a > certain time: > > "Control channel packets shall be processed at the head of the queue, > and shall observe the timestamp semantics described above." > > In theory this enables us to send a control channel packet that tunes() > the USRP to a different frequency and have it executed at the precise > time we want the frequency hop to happen. > > The latency between USRP and host is between 1-30 gsm bursts. We could > send several control channel packets in advance to the USRP so that > the frequency change happens exactly when we need it to. > > How large is the queue on the USRP to store 'delayed control channel packets'? > 64? > > What control channel packet do I have to send to tune() the USRP to > a different frequency? > > regards, > steve
Hi Steve, With the USRP1, it's hard to hop precisely if you've got to steer the VCO on the front end, since depending on the d'board, not all required pins connect to the FPGA (some go only to the FX2). Steering the DDC shouldn't be a problem, but I'm not sure anybody's tested it yet. What daughterboard are you using? FWIW, I merged the latest inband code into the trunk a few days ago. And yes, I really will get the mblock and flow graph code interoperating RSN. Eric _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
