In a direct sampling RF front-end, we can get real samples. But when we sample the signal with two ADC with a sampling delay $ \frac{4n+1}{4fc}$ where $n$ can be $0,1,2 \dots$ Then we have In phase(I) and Quadrature phase(Q) samples outs of two ADC. But there is a problem when we sample "I" before symbol transition and "Q" after transition. This creates a mismatch in correlation with a known signal(local code). As we increase the n value the probability of error samples becomes more.

  • Is there a way to predict the probability of happening?
  • How can I correct the error pairs?
  • $\begingroup$ Is your signal transitioning so fast that this is truly an issue? Usually the bandwidth of the signals is small compared to their carrier frequency and when sampling at more than Nyquist rate I doubt that the transitions are quick enough. $\endgroup$ – jan Sep 17 '13 at 12:21
  • 1
    $\begingroup$ In most direct I/Q sampling receivers I've seen, the ADCs run off of a common clock, so they sample simultaneously. Is this really not the case? One alternative would be to perform some preprocessing on the two sample streams to time-align them first. $\endgroup$ – Jason R Sep 27 '13 at 13:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.