Although CSD coding allows for a minimal quasiternary representation of any binary number, this code has been used in a small number of applications. Surely this is connected with the fact that this code can be generated only recursively from the least-significant to the most-significant bit. As a consequence of this disadvantage, in most applications the nonminimal modified Booth recoding is preferred, because it can be generated in parallel. The Letter introduces a look-ahead method speeding up the generation of the CSD code. By using the CSD code instead of any other quasiternary code, reduced hardware expenditure and reduced power dissipation can be expected due to the minimal number of covered digits.
References
-
-
1)
-
K. Hwang
.
(1979)
Computer arithmetic.
-
2)
-
L. Glasser ,
D. Dobberpuhl
.
(1985)
The design and analysis of VLSI circuits.
-
3)
-
E. Lüder
.
Linear digital signal processing with coefficients represented by sumsof products.
Arch. Elektron. Übertrag.tech.
,
4 ,
229 -
232
-
4)
-
P.R. Moon ,
G.O. Martens
.
A digital filter structure requiring only m-bit delays, shifters, inverters,and m-bit adders plus simple logic circuitry.
IEEE Trans.
,
10 ,
901 -
908
-
5)
-
Hentschke, S., Herrfeld, A., Reifschneider, N., Förster, D., Heinemann, M., Wicke, R.: `A flexible repetitive CSD code filter processing unit in CMOS', Proc. 7th Annual IEEE Int. ASIC Conf. and Exhibit., 1994, p. 261–264.
-
6)
-
G.W. Reitwiesner
.
(1960)
Binary arithmetic, Advances in computers.
-
7)
-
H. Samueli
.
An improved search algorithm for the design of multiplierless FIR filterswith powers-of-two coefficients.
IEEE Trans.
,
7 ,
1044 -
1047
-
8)
-
C.K. Koc ,
S. Johnson
.
Multiplication of signed-digit numbers.
Electron. Lett.
,
11 ,
840 -
841
http://iet.metastore.ingenta.com/content/journals/10.1049/el_19950334
Related content
content/journals/10.1049/el_19950334
pub_keyword,iet_inspecKeyword,pub_concept
6
6