I, and I expect others too, are reluctant to "help" with such "obvious problems" (pretty much your words). You may understand the issues, but you don't want to analyze what your problem is by yourself. Never good. So, let's try the "feed a village" approach: You seem to have good program structure, and you say you have done everything "by the book" but it doesn't work. So, what could be wrong? Dah. There's a mistake in there somewhere, either in execution or in understanding the theory or matlab principles. Begin by reviewing all that.

My suggestion: break down the problem into levels/steps and verify each one-by-one. Do what circuit designers do sometimes and include "test points" in your programs that conditionally output test data (i.e. graph it). If that data's right and as expected (although, your expectations may be wrong), then disable the debug output and advance to the next step. If the output is not as expected, then double check your expectations. If any step fails, then the error must be within the preceding step or your expectation of the previous output. It may be necessary to divide the problem into separate routines for verification.

These are some of the tools engineers use which should have been instilled into you by good teachers.

Now good luck and good hunting.