Title :
Improved minimum-order augmented pipelining algorithm
Author :
Chang, KyungMi ; Lin, XuDuan
Author_Institution :
Electron. & Telecommun. Res. Inst., Taejon, South Korea
Abstract :
Firstly, a comparison of overall performance between minimum-order augmented pipelining (MAP) clustered look-ahead (CLA) filters and scattered look-ahead (SLA) filters is extensively achieved. From the results, the MAP algorithm is revealed to have a problematic domain searching procedure for finding optimized pipelining coefficients. To solve the problems and improve the numerical performance, we propose an improved MAP (IMAP) algorithm, which is especially beneficial for high-Q ultra-high-speed digital filters. The IMAP algorithm is optimized in the aspect of the minimization of augmented pipelining order and undesirable quantization effects simultaneously. Performance of IMAP is compared with that of the conventional MAP algorithm, to demonstrate the advantage of the IMAP algorithm
Keywords :
digital filters; filtering theory; minimisation; noise; pipeline processing; quantisation (signal); recursive filters; MAP algorithm; clustered look-ahead filters; high-Q ultra-high-speed digital filters; minimization; minimum-order augmented pipelining algorithm; numerical performance; optimized pipelining coefficients; quantization effects; scattered look-ahead filters; Clustering algorithms; Delay effects; Digital filters; Minimization methods; Parallel processing; Pipeline processing; Prototypes; Scattering; Transfer functions; Very large scale integration;
Conference_Titel :
VLSI Signal Processing, VIII, 1995. IEEE Signal Processing Society [Workshop on]
Conference_Location :
Sakai
Print_ISBN :
0-7803-2612-1
DOI :
10.1109/VLSISP.1995.527520
