Optimizing Cosine Window Functions by Differential Evolution

This paper proposed an evolutionary method to optimizing the cosine window family by using Differential Evolution (DE). First, the classic cosine windows such as Hamming window and Blackman window are optimized. Results show better performance in terms of side-lobe level and roll-off ratio for the same main-lobe width. Then, a series of optimized six order cosine windows are provided and their performances are analyzed. The comparison proves that the proposed method is an effective approach for the design of window functions.