Abstract
The spectral correlation density (SCD) function is the time-averaged correlation of two spectral components used for analyzing 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 article, 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 1,054 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 4,832 MOPS/W, which is 20.6× better than the published state-of-the-art GPU implementation. In terms of throughput, it achieves 15,340 windows/s (15,340 windows/s × 2,048 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.
- [1] . 2007. Cyclic spectral analysis in practice. Mechan. Syst. Sig. Process. 21, 2 (2007), 597–630.Google Scholar
Cross Ref
- [2] . 2009. Cyclostationarity by examples. Mechan. Syst. Sig. Process. 23, 4 (2009), 987–1036.Google Scholar
Cross Ref
- [3] . 2012. Detection of surface ships from interception of cyclostationary signature with the cyclic modulation coherence. IEEE J. Ocean. Eng. 37, 3 (2012), 478–493.Google Scholar
Cross Ref
- [4] . 2017. Fast computation of the spectral correlation. Mechan. Syst. Sig. Process. 92 (2017), 248–277.Google Scholar
Cross Ref
- [5] . 2015. GPU and FPGA based architecture design for real-time signal classification. In Proceedings of the Wireless Innovation Forum Conference on Wireless Communications Technologies and Software Defined Radio (WInnComm’15). Springer, 70–79.Google Scholar
- [6] . 2016. Reducing memory requirements for high-performance and numerically stable Gaussian elimination. In Proceedings of the ACM/SIGDA International Symposium on Field-programmable Gate Arrays. 244–253.Google Scholar
Digital Library
- [7] . 2018. A faster algorithm for the calculation of the fast spectral correlation. Mechan. Syst. Sig. Process. 111 (2018), 113–118.Google Scholar
Cross Ref
- [8] . 1993. Digital implementations of spectral correlation analyzers. IEEE Trans. Sig. Process. 41, 2 (1993), 703–720.Google Scholar
Digital Library
- [9] . 1996. Detection and Identification of Cyclostationary Signals.
Technical Report . Naval Postgraduate School, Monterey, CA.Google Scholar - [10] . 1986. The spectral correlation theory of cyclostationary time-series. Sig. Process. 11, 1 (1986), 13–36.Google Scholar
Digital Library
- [11] . 1989. Statistical Spectral Analysis: A Nonprobabilistic Theory. Prentice-Hall, Englewood Cliffs, NJ.Google Scholar
- [12] . 1994. Cyclostationarity in Communications and Signal Processing. IEEE Press, New York.Google Scholar
- [13] . 2006. Cyclostationarity: Half a century of research. Sig. Process. 86, 4 (2006), 639–697.Google Scholar
Digital Library
- [14] . 2008. A parallel computing based spectrum sensing approach for signal detection under conditions of low SNR and Rayleigh multipath fading. In Proceedings of the 3rd IEEE Symposium on New Frontiers in Dynamic Spectrum Access Networks. IEEE, 1–10.Google Scholar
Cross Ref
- [15] 2018. RF Datasets for Machine Learning. Retrieved from https://www.deepsig.ai/datasets.Google Scholar
- [16] 2021. UltraScale Architecture DSP Slice User Guide. Retrieved from https://docs.xilinx.com/v/u/en-US/ug579-ultrascale-dsp.Google Scholar
- [17] . 1985. VLSI array processors. IEEE ASSP Magazine 2, 3 (1985), 4–22.Google Scholar
- [18] . 2012. Parallelization of spectrum sensing algorithms using graphic processing units. In Proceedings of the Cross Strait Quad-Regional Radio Wireless Conference (CSQRWC’12). IEEE, 35–39.Google Scholar
- [19] . 2003. Design of a parameterizable silicon intellectual property core for QR-based RLS filtering. IEEE Trans. Very Large Scale Integ. Syst. 11, 4 (2003), 659–678.Google Scholar
Digital Library
- [20] . 2018. GPU based quarter spectral correlation density function. In Proceedings of the Conference on Design and Architectures for Signal and Image Processing (DASIP’18). IEEE, 88–93.Google Scholar
Cross Ref
- [21] . 2020. GPGPU-based parallel implementation of spectral correlation density function. J. Sig. Process. Syst. 92, 1 (2020), 71–93.Google Scholar
Digital Library
- [22] . 2014. Spectrum sensing based on spectral correlation for cognitive radio systems. In Proceedings of the 4th International Conference on Wireless Communications, Vehicular Technology, Information Theory and Aerospace & Electronic Systems (VITAE’14). IEEE, 1–4.Google Scholar
Cross Ref
- [23] . 1996. VLSI systolic arrays for adaptive nulling [radar]. IEEE Sig. Process. Mag. 13, 4 (1996), 29–49.Google Scholar
Cross Ref
- [24] . 2009. Automatic modulation classification for cognitive radios using cyclic feature detection. IEEE Circ. Syst. Mag. 9, 2 (2009), 27–45.Google Scholar
Digital Library
- [25] . 1991. Computationally efficient algorithms for cyclic spectral analysis. IEEE Sig. Process. Mag. 8, 2 (1991), 38–49.Google Scholar
Cross Ref
- [26] . 2009. Identification and Classification of OFDM Based Signals Using Preamble Correlation and Cyclostationary Feature Extraction.
Technical Report . Naval Postgraduate School, Monterey, CA.Google Scholar - [27] . 1999. The strip spectral correlation algorithm for spectral correlation estimation of digitally modulated signals. In 4th International Conference on Telecommunications in Modern Satellite, Cable and Broadcasting Services. TELSIKS’99 (Cat. No. 99EX365), Vol. 1. IEEE, 277–280.Google Scholar
Cross Ref
- [28] . 2020. Real-time automatic modulation classification using RFSoC. In Proceedings of the International Parallel and Distributed Processing Symposium Workshops (IPDPSW’20). IEEE, 82–89.Google Scholar
Cross Ref
- [29] . 2010. Parallelized cyclostationary feature detection on a software defined radio processor. In Proceedings of the International Symposium on Signals, Systems and Electronics. IEEE, 1–4.Google Scholar
Cross Ref
- [30] . 2017. Automated systolic array architecture synthesis for high throughput CNN inference on FPGAs. In Proceedings of the 54th Annual Design Automation Conference. 1–6.Google Scholar
Digital Library
- [31] . 1967. The use of fast Fourier transform for the estimation of power spectra: A method based on time averaging over short, modified periodograms. IEEE Trans. Aud. Electroacoust. 15, 2 (1967), 70–73.Google Scholar
Cross Ref
- [32] . 2020. The RFML ecosystem: A look at the unique challenges of applying deep learning to radio frequency applications. arXiv preprint arXiv:2010.00432 (2020).Google Scholar
- [33] . 2012. Portable and scalable FPGA-based acceleration of a direct linear system solver. ACM Trans. Reconfig. Technol. Syst. 5, 1 (2012), 1–26.Google Scholar
Digital Library
Index Terms
A Scalable Systolic Accelerator for Estimation of the Spectral Correlation Density Function and Its FPGA Implementation
Recommendations
FPGA-based configurable systolic architecture for window-based image processing
Image processing requires more computational power and data throughput than most conventional processors can provide. Designing specific hardware can improve execution time and achieve better performance per unit of silicon area. A field-programmable-...
An FPGA implementation for neural networks with the FDFM processor core approach
This paper presents a field programmable gate array FPGA implementation of a three-layer perceptron using the few DSP blocks and few block RAMs FDFM approach implemented in the Xilinx Virtex-6 family FPGA. In the FDFM approach, multiple processor cores ...
A hardware-efficient computing engine for FPGA-based deep convolutional neural network accelerator
AbstractDeep convolutional neural networks (DCNNs) have recently emerged as a promising approach for computer vision tasks with many new DCNN architectures proposed to further improve their performance. However, the significant computation ...






Comments