We are planning to use the AD9361 for a BPSK based CDMA system. In this mode our BPSK modulator outputs the modulation symbols on I(In-phase) channel only. We set the Q(Quadrature Phase) channel to zeros. Is this acceptable.
We had this CDMA based BPSK implemented on the FMCOMMS2 board. Please see the attached screen shots of the spectrum Analyser. We have mapped the signal as following . If the BPSK modulation symbol is -1 = (I=-1, Q=0) and 1 = (I=1, Q=0). We have applied suitable scaling before sending.However when we captured the signal at Rx side, we found that the amplitude of I channel to be varying. Please see the attached screen shots.
We had set the AGC gain control mode to manual and set the gain to 0dB. The carrier frequency used was 2.4 GHz.
Setting Q channel should not be a problem to set to zero.
Could timing possibly be an issue? Do you get a similar result if you keep I=0? Have you tried another evaluation platform?
We are using the IIO Streaming code for this experimentation. The sampling rate used is 2.4384 MHz. The entire Tx data is stored in a buffer and is played out . The Rx capture is done using the same code. I have attached the IQ data.
Could you please elaborate more on timing . We have tried on another evaluation platform and found same observations.
The problem is very trivial. Once you convert into RF with one card and then back to I-Q with another card the following effects add to the signal.
1. Carrier offset
2. I-Q imbalance.
3. Sampling instant offset.
If your receiver doesn't have blocks to remove these impairments you will see things like what you have shown in the jpg files. When you do this playback game at I-Q level the impairments won't come into picture. So it has nothing to do with AD9361. Also for AD9361 please put some gain as per the input signal strength so that the signal is well above the ADC's noise floor.
Retrieving data ...