My interest area is: modern control system; robust controller design for uncertain systems; smart structure modelling; multirate controller; DSP algorithm; model order reduction; softcomputing; MOR of time interval system.
Step 1: Pick a problem that you care about and that you think you can solve using the areas of knowledge you listed (even if you don't at the moment have sufficient knowledge of these). Step 2: Do a literature search to see what people have done on that problem. If it hasn't been solved to your satisfaction, and you think you can contribute, pick it as your PhD topic. If it has been solved, or you find it uninteresting, go back to Step 1.:)