Partial Reconfigurable FIR Filtering System Using Distributed Arithmetic. (6th December 2010)

• Record Type: Journal Article
• Authors: Llamocca, Daniel; Pattichis, Marios; Vera, G. Alonzo
• Other Names: Prasanna Viktor K. Academic Editor.
• Is Part Of: International journal of reconfigurable computing. Volume 2010(2010)
• Publication Date: 2010-12-06
• Subjects: Adaptive computing systems -- Periodicals; Adaptive computing systems; Periodicals; 004
• DOI: 10.1155/2010/357978
• Languages: English
• ISSNs: 1687-7195

Abstract

Abstract : Dynamic partial reconfiguration (DPR) allows us to adapt hardware resources to meet time-varying requirements in power, resources, or performance. In this paper, we present two new DPR systems that allow for efficient implementations of 1D FIR filters on modern FPGA devices. To minimize the required partial reconfiguration region (PRR), both implementations are based on distributed arithmetic. For a smaller required PRR, the first system only allows changes to the filter coefficient values while keeping the rest of the architecture fixed. The second DPR system allows full FIR-filter reconfiguration while requiring a larger PR region. We investigate the proposed system performance in terms of the dynamic reconfiguration rates. At low reconfiguration rates, the DPR systems can maintain much higher throughputs. We also present an example that demonstrates that the system can maintain a throughput of 10 Mega-samples per second while fully reconfiguring about seventy times per second.