Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Coherence between two signals

  1. Jun 29, 2013 #1
    The coherence I'm asking about is also known as magnitude-squared coherence.

    Let's say we have input A(t) and we look at the output B(t), the system L is linear but it has damping effect on the signals. In a long time, this damping will literally kill the high frequencies to zero.

    My question is if A(t) contains a broadband spectrum, and let A(t) go through the system L, will the coherence between A(t) and B(t) be less than 1, or just equal to 1?

    Jo
     
  2. jcsd
  3. Jun 29, 2013 #2

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor
    2016 Award

    If I understand you correctly: if L is a deterministic linear system then the coherence properties of B(t) should be equal to the coherence properties of A(t). IIRC, the transfer function of L determines how the coherence propagates from A(t) to B(t), and while stochastic systems will degrade signal coherence, deterministic systems do not.

    There is some discussion here: http://www.dtic.mil/dtic/tr/fulltext/u2/682486.pdf
     
  4. Jun 30, 2013 #3
    Hi Andy,

    Thanks for your reply.

    OK, I see what you mean. By stochastic system, do you mean there is noise going into the linear system? Of course, in this case the coherence would decay.

    Ley's consider the damping being very strong, literally killing all the high frequencies. Let's say we have A(t)=sin(2*pi*t)+sin(2*pi*9*t) as an input. Going through the linear system with damping, the high frequency 9 Hz is killed, so B(t)=c*sin(2*pi*t), c is some constant less than 1 because frequency 1 Hz is also damped. Do you mean the coherence between A(t) and B(t) is 1 for frequency 9Hz? I can't see that.

    Jo
     
  5. Jul 2, 2013 #4

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor
    2016 Award

    Not exactly- although noise is often modeled with stochastic equations. 'Stochastic' simply means that the time evolution of a system is not deterministic- to model a stochastic processes (e.g. diffusion- Brownian motion is the 'canonical problem') requires statistical analysis.

    I don't really understand how you are modeling the damping- but in any case, if you compute <A(t)A(t+τ)>/<A^2> and compare that to <B(t)B(t+τ)>/<B^2>, you will have your answer.

    A possible reference for you is Marathay's "Elements of Optical Coherence Theory", which is more general than you need but is very complete.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook