Triple based number systems - a novel concept for performance enhancement of digital signal processors

The signal processing algorithms face many challenges in real-time applications because of their high computational complexity. Therefore, the major issues have been the enhancement of speed of the arithmetic units in general and multiplications and additions in particular. Double based number systems (DBNS) are increasingly gaining popularity for their capabilities of handling arithmetic operations efficiently. Even though DBNS schemes exhibit reasonably good performance for 8 bit multiplication, they are not efficient for higher bits .In this paper we introduce a new concept ldquotriple based number systems (TBNS) for performance enhancement of the multiplier of the digital signal Processors. Efficiency of this number system has been dealt with in details and a comparison between TBNS and DBNS clearly indicate the advantages of the former in terms of speed, hardware complexity and power dissipation. Different architectural models have been proposed. Finally, an architecture of high precision FIR filter has been presented which exploits the advantages of the TBNS scheme.

[1]  Keshab K. Parhi,et al.  VLSI digital signal processing systems , 1999 .

[2]  W.C. Miller,et al.  On efficient techniques for difficult operations in one and two-digit DBNS index calculus , 2000, Conference Record of the Thirty-Fourth Asilomar Conference on Signals, Systems and Computers (Cat. No.00CH37154).

[3]  Markus Weinhardt,et al.  PACT XPP—A Self-Reconfigurable Data Processing Architecture , 2004, The Journal of Supercomputing.

[4]  Roberto Muscedere,et al.  Efficient techniques for binary-to-multidigit multidimensional logarithmic number system conversion using range-addressable look-up tables , 2005, IEEE Transactions on Computers.

[5]  Ketan Mayer-Patel,et al.  Exploiting temporal parallelism for software-only video effects processing , 1998, MULTIMEDIA '98.

[6]  Graham A. Jullien,et al.  Theory and applications for a double-base number system , 1997, Proceedings 13th IEEE Sympsoium on Computer Arithmetic.

[7]  B. Lee.  A new algorithm to compute the discrete cosine Transform , 1984 .

[8]  Majid Ahmadi,et al.  A hybrid DBNS processor for DSP computation , 1999, ISCAS'99. Proceedings of the 1999 IEEE International Symposium on Circuits and Systems VLSI (Cat. No.99CH36349).

[9]  Quentin F. Stout,et al.  Reconfigurable SIMD massively parallel computers , 1991 .

[10]  Graham A. Jullien,et al.  Near canonic double-based number system (DBNS) with applications in digital signal processing , 1996, Optics & Photonics.

[11]  Russell Tessier,et al.  c ○ 2001 Kluwer Academic Publishers. Manufactured in The Netherlands. Reconfigurable Computing for Digital Signal Processing: A Survey ∗ , 1999 .

[12]  Keshab K. Parhi,et al.  VLSI digital signal processing systems , 1999 .

[13]  Benjamin W. Wah,et al.  The Design of Optimal Systolic Arrays , 1985, IEEE Transactions on Computers.

[14]  Roberto Muscedere,et al.  A 2-digit DBNS filter architecture , 2000, 2000 IEEE Workshop on SiGNAL PROCESSING SYSTEMS. SiPS 2000. Design and Implementation (Cat. No.00TH8528).

[15]  Howard Jay Siegel,et al.  PASM: A Partitionable SIMD/MIMD System for Image Processing and Pattern Recognition , 1981, IEEE Transactions on Computers.