Convex analysis and global optimization of output convariance constraint problems

This paper addresses the problem of designing state or observer-based output feedback controllers for continuous time linear system, subject to output covariance matrix constraints. In a case when the state is available for feedback, the above problem can also be solved taking into account parameter uncertainties which are assumed to belong to convex bounded domains. The basic problem under consideration is converted into a convex programming problem which is solved by two different but equivalent procedures (primal and dual versions). The numeric behavior of both procedures are compared with those available in the literature to date