Summary
Let M(N) denote the number of active multiplications/divisions needed to compute the discrete Fourier transform of N variables. Winograd showed M(N) ≦ O(N). For N a power of two we analyse his method and find M(N)≦8 N−o(N). Using additional symmetries we improve this to M(N)≦2N−o(N). We also give a very short proof for M(N)≦4N-o(N). Finally we show M(N)≧N − 2 log N.
Similar content being viewed by others
References
Artin, E.: Galoissche Theorie. DTB Nr. 21, Zürich 1973
Bluestein, L.I.: A linear filter approach to the computation of the discrete Fourier transform. Necem Rec. 218–219 (1968)
Borodin, A., Munro, I.: The computational complexity of algebraic and numeric problems. American Elsevier, 1975
Cooley, J.W., Tukey, J.W.: An algorithm for the machine calculation of complex Fourier series. Math. Comput. 19, 297–301 (1965)
Reiffen, HJ., Scheja, G., Vetter, U.: Algebra. BI-HTB Bd. 110, Mannheim 1969
Winograd, S.: On the number of multiplications necessary to compute certain functions. Comm. Pure Appl. Math. 23, 165–179 (1970)
Winograd, S.: The effect of the fild of constants on the number of multiplications. Proceedings of the 16th Annual Symposium on Foundations of Computer Science, 1975, pp. 1–2
Winograd, S.: On computing the discrete Fourier transform. Proc. Nat. Acad. Sci. 73, 4, 1005–1006 (1976)
Tarjan, R.E.: Complexity of combinatorial algorithms. Technical report. Stanford University STAN-CS, 77–609, April 1977
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Mescheder, B. On the number of active *-operations needed to compute the discrete Fourier transform. Acta Informatica 13, 383–408 (1980). https://doi.org/10.1007/BF00288772
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF00288772