GPU Based Quarter Spectral Correlation Density Function

Marshall, Scott; Vanhoy, Garrett; Akoglu, Ali; Bose, T.; Ryu, Bo

doi:10.1109/dasip.2018.8596977

Cited by 4 publications

(1 citation statement)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lee et al [18] developed a GPU implementation of the FAM method and achieved a speedup of 39× over the serial implementation running on a 2.94 GHz Intel Core 2 CPU. The University of Arizona proposed high throughput GPU implementations [20,21] and a heterogeneous FPGA-GPU implementation [5] of FAM. While they reported a state-of-the-art throughput of 3300 windows/s on a Tesla K40 GPU, the hybrid FPGA-GPU implementation was signiicantly less optimised, achieving only 19 windows/s.…”

Section: Existing Implementations Of the Fam Methodsmentioning

confidence: 99%

A Scalable Systolic Accelerator for Estimation of the Spectral Correlation Density Function and Its FPGA Implementation

Maskell

et al. 2022

ACM Trans. Reconfigurable Technol. Syst.

View full text Add to dashboard Cite

The spectral correlation density (SCD) function is the time-averaged correlation of two spectral components, used for analysing periodic signals with time-varying spectral content. Although the analysis is extremely powerful, it has not been widely adopted in real-time applications due to its high computational complexity. In this paper, we present an efficient FPGA implementation of the FFT accumulation method (FAM) for estimating the SCD function and its alpha profile. The implementation uses a linear systolic array with a bi-directional datapath consisting of DSP-based processing elements (PEs) with a dedicated instruction schedule, achieving a PE utilization of 88.2%. The 128-PE implementation achieves a clock frequency in excess of 530 MHz and consumes 151K LUTs, 151K FFs, 264 BRAMs, 4 URAMs and 1054 DSPs, which is less than 36% of the logic fabric on a Zynq UltraScale+ XCZU28DR-2FFVG1517E RFSoC device. It has a modest 12.5W power consumption and an energy efficiency of 4832 MOPS/W which is 20.6 × better than the published state-of-the-art GPU implementation. In terms of throughput, it achieves 15340 windows/s (15340 windows/s × 2048 samples/window = 31.4 MS/s), which is a 4.65 × improvement compared to the above-mentioned GPU implementation and 807 × compared to an existing hybrid FPGA-GPU implementation.

show abstract

Section: Existing Implementations Of the Fam Methodsmentioning

confidence: 99%