Dan, I sampled an UTRA FDD downlink signal and implemented the widely know standard cell search algorithm (see for example http://www.umtsworld.com/technology/cell_search.htm). To understand the details, you need to download the right document in the standard (ETSI TS 125 113, if I remember correctly) which is available at www.etsi.com. There are also loads of papers at the IEEE.
The algorithm: 0. Resampling to N*3.84 Mcps, internal oversampling needed to compensate for non-ideal sampling, matched filtering 1. DFT stage for coarse frequency aquisition using the PSC (synch to slot) 2. correlation with SSCs to get code group (synch to frame) 3. try the 8 different scrambling codes in code group 4. fine frequency aquisition using known symbols on the CPICH 5. start to despread data channels using CPICH as phase reference Nothing fancy at all. If it works in Matlab, it should work this way in GNU Radio, too? Would be a nice project, I guess. I'm not planning on implementing the whole show in Matlab (which would be close to impossible for me). One purpose was to show that this works with the USRP. A tricky part is the scrambling code generation, which is described not very clearly in the standard. If you have further questions, just ask. Jens On Sat, Jun 10, 2006 at 09:24:19AM -0400, Dan Veeneman wrote: > Hi Jens, > > I'm quite interested in WCDMA, and spread spectrum in general, and > would certainly be interested in learning more about your algorithm. > If you can't send the code, do you have a description of the final > algorithm you worked out? Perhaps even a small section of the > code that does the search and despreading? > > > Regards, > > Dan > > > At 04:12 AM 6/10/06, you wrote: > >Hello everybody, > > > >I did some work the last couple of weeks implementing the cell search > >algorithm and despreading of basic channels for WCDMA (UMTS) using the USRP > >in Matlab. > > > >Unfortunately I do not have the copyright, so I can't post the code. > > > >But if anyone is interested in playing with WCDMA in GNU Radio, I'd be > >glad to share my insight. For now I'll get back to DAB. > > > >Jens > > > > > >_______________________________________________ > >Discuss-gnuradio mailing list > >[email protected] > >http://lists.gnu.org/mailman/listinfo/discuss-gnuradio > _______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
