Application Specific Processors pp 161-195 | Cite as

# Multidimensional Systolic Arrays for Computing Discrete Fourier Transforms and Discrete Cosine Transforms

## Abstract

This chapter presents a new approach for computing the multidimensional discrete Fourier transform (DFT) and the multidimensional discrete cosine transform (DCT) in a multidimensional systolic array. There are extensive applications of fast Fourier transform (FFT) and fast cosine transform (FDCT) algorithms. From the basic principle of fast transform algorithms (breaking the computation in successively smaller computations), we find that the multidimensional systolic architecture is efficiently used for implementing FFT algorithms and FDCT algorithms. The essence of the multidimensional systolic array is to combine different types of semi-systolic arrays into one array so that the resulting array becomes truly systolic. This systolic array does not require any preloading of input data and it generates output data only from boundary PEs. No networks for transposition between intermediate constituent 1-D transforms are required; therefore the entire processing is fully pipelined. This approach is well suited for VLSI implementation by providing simple and regular structures. Complexity estimation of *area***time* ^{2} shows our multidimensional systolic array is within a factor of *logk* of the lower bound for an *M*-dimensional *k*-point DFT (*k*=*N* ^{ M }).

## Keywords

Discrete Cosine Transform Discrete Fourier Transform Processing Element Systolic Array Chinese Remainder Theorem## Preview

Unable to display preview. Download preview PDF.

## References

- 6-1.A. V. Oppenheim and R. W. Schafer,
*Discrete-Time Signal Processing*, Englewood Cliffs, NJ: Prentice-Hall, 1989.MATHGoogle Scholar - 6-2.I. Gertner and M. Shamash, “VLSI Architectures for Multidimensional Fourier Transform Processing,”
*IEEE Trans. on Computers*, Vol. C-36, 1987, pp. 1265–1274.CrossRefGoogle Scholar - 6-3.S. Y. Kung,
*VLSI Array Processors*, Englewood Cliffs, NJ: Prentice-Hall, 1988.Google Scholar - 6-4.H. T. Kung, “Memory Requirements for Balanced Computer Architecture,”
*Proc. 13th Annual Int. Symp. on Computer Architecture*, June 1986, pp. 49–54.Google Scholar - 6-5.N. R. Murthy and M. N. S. Swamy, “On die Real-Time Computation of DFT and DCT through Systolic Architectures”,
*IEEE Trans, on Signal Processing*, Vol. 42, Apr. 1994, pp. 988–991.CrossRefGoogle Scholar - 6-6.E. E. Swartzlander, “Systolic FFT Processors,” W. Moore, A. McCabe, and R. Urquhart, eds.,
*Proceedings International Workshop on Systolic Arrays*, Boston: Adam Hilger, 1987, pp. 133–140.Google Scholar - 6-7.E. E. Swartzlander, ed.,
*Systolic Signal Processing Systems*, New York: Marcel Dekker, Inc., 1987.Google Scholar - 6-8.T. D. Roziner and M.G. Karpovsky, “Multidimensional Fourier Transforms by Systolic Architectures,”
*Journal of VLSI Signal Processing*, 4, 1992, pp. 343–354.CrossRefGoogle Scholar - 6-9.M. H. Lee, “High Speed Multidimensional Systolic Arrays for Discrete Fourier Transform,”
*IEEE Trans. Circuits and Systems-II: Analog and Digital Signal Processing*, Vol. 39, 1992, pp. 876–879.MATHCrossRefGoogle Scholar - 6-10.H. S. Lim, “A Study on die Parallel Processing Architectures for 2-D Discrete Cosine Transform,”
*M. S. Thesis*, Ch. 3, Seoul National University, Korea, 1991.Google Scholar - 6-11.H. S. Lim and Earl E. Swartzlander, Jr., “A Systolic Array for 2-D DFT and 2-D DCT,”
*International Conference on Application-Specific Array Processors*, Aug. 1994, pp. 123–131.Google Scholar - 6-12.H. S. Lim and Earl E. Swartzlander, Jr., “An Efficient Systolic Array for DCT Based on Prime-Factor Decomposition,”
*International Conference on Computer Design*, Oct. 1995, pp. 644–649.Google Scholar - 6-13.H. S. Lim and Earl E. Swartzlander, Jr., “Efficient Systolic Arrays for FFT Algorithms,” 29
^{th}*Annual Asilomar Conference on Signals, Systems, and Computers*, Oct. 1995, pp.141–145.Google Scholar - 6-14.H. S. Lim and Earl E. Swartzlander, Jr., “Multidimensional Systolic Arrays for Multidimensional DFTs,”
*IEEE International Conference on Acoustics, Speech & Signal Processing*, May 1996, pp. 3277–3280.Google Scholar - 6-15.H. S. Lim, C. Yim and Earl E. Swartzlander, Jr., “Finite Word-Length Effects of a 2-D DCT Systolic Array,”
*International Conference on Application-Specific Array Processors*, Aug. 1996.Google Scholar - 6-16.N. Ling and M.A. Bayoumi, “An Algoridim Transformation Technique for Multi-Dimensional DSP Systolic Arrays,”
*IEEE International Symposium on Circuits and Systems*, 1988, pp. 2275–2278.Google Scholar - 6-17.C. N. Zhang and D.Y.Y. Yun, “Multi-Dimensional Systolic Networks for Discrete Fourier Transform,”
*Proc. 11*^{th}*Int. Symp. Computer Architecture*, Ann Arbor, Mich., 1984, pp. 215–222.Google Scholar - 6-18.G. H. Allen, P.B. Denyer, and D. Renshaw, “A Bit Serial Linear Array DFT,”
*IEEE International Conference on Acoustics, Speech & Signal Processing*, San Diego, 1988, pp. 41A.1.1–41A.1.4.Google Scholar - 6-19.M.A. Bayoumi, G.A. Jullien, and W.C. Miller, “A VLSI Array for Computing the DFT based on RNS,”
*IEEE International Conference on Acoustics, Speech & Signal Processing*, Tokyo, 1986, pp. 2147–2150.Google Scholar - 6-20.J. A. Beraldin, T. Aboulnasr, and W. Steenaart, “Efficient One-Dimensional Systolic Array Realization of die Discrete Fourier Transform,”
*IEEE Trans. Circuits and Systems*, Vol. 36, 1989, pp. 95–100.CrossRefGoogle Scholar - 6-21.L.W. Chang and M.Y. Chen, “A New Systolic Array for Discrete Fourier Transform,”
*IEEE Trans. Acoustics, Speech, and Signal Processing*, Vol. ASSP-36, 1988, pp. 1665–1666.CrossRefGoogle Scholar - 6-22.J. Guo, C. Liu, and C. Jen, “The Efficient Memory-based VLSI Array Designs for DFT and DCT,”
*IEEE Trans, on Circuits and Systems-II: Analog and Digital Signal Processing*, Vol. 39, 1992, pp. 723–733.MATHCrossRefGoogle Scholar - 6-23.D. C. Kar and V.V. Rao, “A New Systolic Realization for Discrete Fourier Transform,”
*IEEE Trans, on Signal Processing*, Vol. 41, 1993, pp. 2008–2010.MATHCrossRefGoogle Scholar - 6-24.B. G. Lee, “A New Algorithm to Compute die Discrete Cosine Transform,”
*IEEE Trans. Acoustics, Speech, and Signal Processing*, Vol. ASSP-32, 1984, pp. 1243–1245.Google Scholar - 6-25.C. Chakrabarti and J. Ja’Ja’, “Systolic Architectures for the Computation of the Discrete Hartley and die Discrete Cosine Transforms Based on Prime Factor Decomposition,”
*IEEE Trans, on Computers*, Vol. 39, Nov. 1990, pp. 1359–1368.MathSciNetCrossRefGoogle Scholar - 6-26.B. G. Lee, “Input and Output Index Mappings for a Prime-Factor-Decomposed Computation of Discrete Cosine Transform,”
*IEEE Trans, on Acoust., Speech, and Signal Processing*, Vol. 37, Feb. 1989, pp. 237–244.MATHCrossRefGoogle Scholar - 6-27.J. Canaris, “A VLSI Architecture for die Real Time Computation of Discrete Trigonometric Transforms,”
*Journal of VLSI Signal Processing*, Vol. 5, 1993, pp. 95–104.MATHCrossRefGoogle Scholar - 6-28.M. Sheu, J. Lee, J. Wang, A. Suen, and L. Liu, “A High Throughput-rate Architecture for 8*8 2-D DCT,”
*IEEE International Symposium on Circuits and Systems*, Vol. 3, 1993, pp. 1587–1590.Google Scholar - 6-29.N. I. Cho and S. U. Lee, “DCT Algorithms for VLSI Parallel Implementation,”
*IEEE Trans. Acoustics, Speech, and Signal Processing*, Vol. ASSP-38, 1990, pp. 121–127.Google Scholar - 6-30.C. D. Thompson, “Fourier Transforms in VLSI,”
*IEEE Trans, on Computers*, Vol. C-32, 1983, pp. 1047–1057.CrossRefGoogle Scholar - 6-31.A. D. Boodi, “A Signed Binary Multiplication Technique,”
*Quarterly Journal of Mechanics and Applied Mathematics*, Vol. 4, 1951, pp. 236–240.MathSciNetCrossRefGoogle Scholar - 6-32.S. Magar, S. Shen, G. Luikuo, M. Fleming, and R. Agular, “An Application Specific Chipset for 100 MHz data rate,”
*IEEE International Conference on Acoustics, Speech & Signal Processing*, 1988, pp.1989–1992.Google Scholar - 6-33.J. O’Brien, J. Mather, and B. Holland, “A 200 MIPS Single-Chip IK FFT Processor,”
*Proc. 1989 IEEE Int. Solid-State Circuits Conf*, 1989, pp. 166–167.Google Scholar - 6-34.R. M. Owens and J. Ja’Ja, “A VLSI Chip for the Winograd/Prime Factor Algorithm to Compute die Discrete Fourier Transform,”
*IEEE Trans, on Acoust., Speech, and Signal Processing*, Vol. 34, 1986, pp. 979–989.CrossRefGoogle Scholar - 6-35.T. K. Truong, I. S. Reed, I. S. Hsu, H. C. Shyu, and H. M. Shao, “A Pipeline Design of a Fast Prime Factor DFT on a Finite Field,”
*IEEE Trans, on Computers*, Vol. 37, 1988, pp. 266–273.MathSciNetMATHCrossRefGoogle Scholar - 6-36.C. S. Burrus, “Index Mappings for Multidimensional Formulation of the DFT and Convolution,”
*IEEE Trans. AcoustSpeech, and Signal Processings*, Vol. 25, 1977, pp. 239–242.MATHCrossRefGoogle Scholar - 6-37.I. Niven and H. S. Zuckennan,
*An Introduction to the Theory of Numbers*, John Wiley & Sons, Third Edition, Ch. 2, 1972.MATHGoogle Scholar - 6-38.P. Z. Lee and F. Y. Huang, “An Efficient Prime-Factor Algoridims for the Discrete Cosine Transform and and Its Hardware Implementations,”
*IEEE Trans, on Signal Processing*, Vol. 42, Aug. 1994, pp. 1996–2005.CrossRefGoogle Scholar - 6-39.P. P. N. Yang and M. J. Narasimha, “Prime Factor Decomposition of the Discrete Cosine Transform and Its Hardware Realization,”
*IEEE International Conference on Acoustics, Speech & Signal Processing*, 1985, pp. 20.5.1–20.5.4.Google Scholar