Search for follwing paper. if you dont get the paper over net send me your personal email id.
my email id is bharat\at\arithos\com
A FAMILY OF COSINE-SUM WINDOWS FOR HIGH-RESOLUTION MEASUREMENTS Hans-Helge Albrecht Physikalisch-Technische Bundesanstalt Abbestra?e 2-12, D-10587 Berlin, Germany Phone: +49 30 3481 311 Fax: +49 30 3481 490 E-mail: hans-helge.albrecht(a)ptb.de
A better terminology might be "minimum sidelobe cosine-sum" window.
harris doesn't seem to have published past 4 terms and his 4 term optimized window had 92 dB sidelobe rejection. Nuttall provided a correction to the 4 term coefficients to yield 98 dB sidelobe rejection. So, harris' 4 term wasn't minimum sidelobe. Nuttall doesn't seem to have published past 4 terms either.
So, if you want a five-term minimum sidelobe cosine-sum window you could use:
A family of cosine-sum windows for high-resolution measurements Albrecht, H.H. Acoustics, Speech, and Signal Processing, 2001. Proceedings. (ICASSP apos;01). 2001 IEEE International Conference on Volume 5, Issue , 2001 Page(s):3081 - 3084 vol.5
The paper gives up to 11 terms and -289 dB sidelobe rejection. At 12 terms there were numerical problems with running the optimization in 64 bit floating point. Albrecht's 4 term coefficients agree with Nuttal's 4 term values. The 7 term window appears in some 24-bit A-D converter application notes published before Albrecht's paper. The values published were consistent with Albrecht's. A number of people have successfully run the optimization for various numbers of coefficients so I go with "minimum sidelobe cosine-sum" window for the family.
Special windows are used in spectral analysis to reduce the effect of spectral leakage. Windows with low sidelobe amplitude are necessary for the detection of small signals when highly dynamic spectra are concerned. The design of a family of cosine-sum windows with minimum sidelobes is described. The coefficients and selected parameters for windows with a peak sidelobe level between 43 dB and 289 dB are stated