MSK timing recovery algorithm.

At the bottom of this page: http://www.mathworks.com/help/comm/ref/msktypesignaltimingrecovery.html

there is a algorithm that for purposes of research I need to reproduce. I'm confused by some of the notation in the equation. Do the parentheses after the r^2 and r*^2 terms refer to the index of the value or something else? What is confusing is the first part kT or (k-1)T seem to refer to the relative index of an item in the input. However Ts and d_k does not.

I have a liscense for the Communications System Toolbox. I just need to understand how the algorithm works. I also have a copy of the document referenced at the bottom of the page and am still confused.

Thank you.

[Jon Boerner]

I believe the parentheses just refer to multiplication. The k does refer to the index of the input, but I believe it just helps to scale the error differently as more symbols are acquired by the feedback loop.

I am not sure if you noticed, but there is also a link at the bottom of the page which describes more completely the algorithm used for the timing recovery:

http://www.mathworks.com/help/comm/ug/synchronization.html#fp62069

The error calculation is just one of the pieces of the feedback loop. The exact implementation of the loop is hidden inside the Simulink blocks, however.

Have you already seen that synchronization page?

[Tyler]

I did see that link. I don't think it's simply multiplication since in the IEEE paper referenced the one that's conjugate is taken and then squared is the sample one symbol period before the current bit. I read the page on the feedback loop and I'm attempting to implement the loop in matlab code without the use of the toolbox functions. I have it working with them now, however that doesn't translate to C where I need it.