High-Throughput FPGA Implementation of Sliding DCT using Look-Ahead Pipelining
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The Discrete Cosine Transform (DCT) is often applied in a sliding window setting. While sliding DCT (SDCT) algorithms take advantage of the computational redundancy inherent from the overlap between consequent processing windows, the recursive computation nature imposes a fundamental limit on realizable throughput. This study presents a high-throughput FPGA implementation of the SDCT algorithm using look-ahead pipelining. A look-ahead transformation of degree two is applied to the original SDCT recursion formula resulting in a threefold reduction of the iteration bound. The resulting maximum clock frequency on a Xilinx FPGA device is increased 2.1x (from 58.4 MHz to 123.4 MHz) by carefully retiming the look-ahead pipeline registers and inserting additional pipeline registers in feed-forward cutsets of the signal-flow graph.
Abstract
The Discrete Cosine Transform (DCT) is often applied in a sliding window setting. While sliding DCT (SDCT) algorithms take advantage of the computational redundancy inherent from the overlap between consequent processing windows, the recursive computation nature imposes a fundamental limit on realizable throughput. This study presents a high-throughput FPGA implementation of the SDCT algorithm using look-ahead pipelining. A look-ahead transformation of degree two is applied to the original SDCT recursion formula resulting in a threefold reduction of the iteration bound. The resulting maximum clock frequency on a Xilinx FPGA device is increased 2.1x (from 58.4 MHz to 123.4 MHz) by carefully retiming the look-ahead pipeline registers and inserting additional pipeline registers in feed-forward cutsets of the signal-flow graph.
참고문헌 (13)
- N. Ahmed, "Discrete cosine transform", IEEE Trans. Comput, C-23(1), pp.90~93 (1974)
- H. Ochoa-Dominguez, "Discrete Cosine Transform" (2019), CRC Press
- 하혜빈, "MV – HEVC Chip Design Using High Level Synthesis", IDEC Journal of Integrated Circuits and Systems, 6(4), pp.23~28 (2020), 반도체설계교육센터[DOI]
- S. Haykin, "Adaptive Filter Theory" (2013), Pearson
- A. V. Oppenheim, "Discrete-time Signal Processing" (2009), Pearson
- M. Puschel, "Algebraic signal processing theory : Cooley–Tukey type algorithms for DCTs and DSTs", IEEE Trans. Signal Process, 56(4), pp.1502~1521 (2008)
- J. Xi, "Computing running DCT's and DST's based on their second-order shift properties", IEEE Trans. Circuits. Syst. I, 47(5), pp.779~783 (2000)
- V. Kober, "Fast algorithms for the computation of sliding discrete sinusoidal transforms", IEEE Trans. Signal Process, 52(6), pp.1704~1710 (2004)
- Q. Li, "Recursive algorithms for direct computation of generalized sliding discrete cosine transforms", Proc. 3rd Int'l. Congress on Image and Signal Process, pp.3017~3020 (2010)
- V. Kober, "Recursive Algorithms for Computing Sliding DCT With Arbitrary Step", IEEE Sensors, 21(10), pp.11507~11513 (2021)
- K. K. Parhi, "VLSI Digital Signal Processing Systems" (1999), Wiley-Interscience
- H.-J. Kang, "FIR filter synthesis algorithms for minimizing the delay and the number of adders", IEEE Trans. Circuits Syst. II Analog Digit. Signal Process, 48(8), pp.770~777 (2001)
- "7 Series FPGAs Data Sheet: Overview" (2020)[URL]