I'm working with a customer for a cell-phone network add-on application.
The application requires that we listen in on cell phones and gather some data about their QOS at certain antennas. My immediate task is trying to de-spread the signal from a cell phone so I can start doing useful work. I'm kind of swimming upstream on this, because when a base station does this task it's set the timing to the phone so it only has to account for the phone's delay; when I do it I'll have to search over a longer time span.
Right now we're trying to collect data from a phone and demodulate it; I'm sifting through a pile of time-domain captures sent to me by the customer, trying to sort them out. I'm finding it particularly frustrating because getting direct-sequence spread spectrum right seems to be like wandering around the desert looking for a hole: you don't know you're close unless you fall into it, so it's hard to know when you have results that are partially correct.
I have TIA/EIA/IS-2000.0-A sitting in front of me. From the document, it appears likely that a CDMA phone will be transmitting on the Access Channel or the Reverse Fundamental Channel during a voice call. Is this correct, or am I cracked?
I had thought, with a single phone, that I could take advantage of the Q-channel PN modulation to at least synchronize the short codes to the phone. This even works with sample vectors that I have generated, but it doesn't work with the real data. I can see how it wouldn't work at all with multiple phones.
Can anyone recommend good material on practical CDMA techniques? I've done this for the customer twice already, with TDMA and with GSM. Both times we just went straight from the standard to reality with very little fuss. This CDMA stuff is proving a tougher nut to crack. I'm looking for a web site or a book, "CDMA demodulation techniques for dummies" is just about the right title for me today.
Even a web site that has samples of the I-channel and Q-channel PN sequence would let me verify that I'm on the right track. Heaven would be a file with a sample from a phone transmitting at a known time on a known channel with a known long code mask. Real code, would, of course, be beyond heaven.
If anyone has a recommendation, I look forward to any help.
Thanks.