Constrained Chebyshev Design of a Class of FIR Filters Using Sequential Unconstrained Optimization Techniques

This paper considers the Chebyshev design of FIR filters with freqwuency inequality constraints. The design problem is formulated as a constrained Chebyshev approximation problem, the solution to which is found by solving a sequence of unconstrained Chebyshev optimization problems. An improved iterative Remez algorithm is proposed and the covergence of the algorithm is obtained. Design examples demonstrate that it is much more efficient than the iterative Remez algorithm proposed earlier by the author and the linear programming based algorithms.