一种伯努利粒子滤波器的FPGA实现

doi:10.12305/j.issn.1001-506X.2025.02.07

系统工程与电子技术 ›› 2025, Vol. 47 ›› Issue (2): 398-405.doi: 10.12305/j.issn.1001-506X.2025.02.07

• 传感器与信号处理 • 上一篇

一种伯努利粒子滤波器的FPGA实现

连红飞, 李东升, 蒋彦雯, 范红旗, 肖怀铁, 王国嫣

国防科技大学电子科学学院自动目标识别全国重点实验室, 湖南长沙 410073

收稿日期:2023-09-25 出版日期:2025-02-25 发布日期:2025-03-18
通讯作者: 范红旗
作者简介:连红飞 (1997—), 男, 博士研究生, 主要研究方向为毫米波雷达阵列成像、嵌入式实时信号处理
李东升 (1994—), 男, 博士研究生, 主要研究方向为多目视觉即时定位与地图创建、目标跟踪
蒋彦雯 (1991—), 女, 副教授, 博士, 主要研究方向为阵列雷达成像、精确制导应用
范红旗 (1978—), 男, 研究员, 博士, 主要研究方向为主动感知系统、目标跟踪、信息融合与智能导引
肖怀铁 (1966—), 男, 教授, 博士, 主要研究方向为自动目标识别、雷达信号处理、机器学习
王国嫣 (1989—), 女, 讲师, 博士, 主要研究方向为激光、多目视觉即时定位与地图创建、多源信息融合、人机交互
基金资助:
国家自然科学基金(62303478)

FPGA implementation of a Bernoulli particle filter

Hongfei LIAN, Dongsheng LI, Yanwen JIANG, Hongqi FAN, Huaitie XIAO, Guoyan WANG

National Key Laboratory of Automatic Target Recognition, College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, China

Received:2023-09-25 Online:2025-02-25 Published:2025-03-18
Contact: Hongqi FAN

摘要/Abstract

摘要：

针对伯努利粒子滤波器在嵌入式应用环境中的高速、高效计算问题, 以雷达微弱目标联合检测估计伯努利粒子滤波器为例, 提出一种功能模块化、粒子规模可扩展的现场可编程门阵列(field programmable gate array, FPGA)实现架构, 并通过粒子状态流水计算、分层累加求和、并行化重采样等手段进一步提高滤波计算速度。Xilinx ZC706评估板板载测试实验证明了所提架构良好的可扩展性和优异的加速比, 当粒子数量为1 024时, 相较于Intel Core i3-4130 CPU计算环境下的加速比约为10⁴量级, 该结果对伯努利粒子滤波技术在雷达、机器人、导航制导等领域的应用具有重要参考价值。

关键词: 伯努利粒子滤波器, 现场可编程门阵列, 实时信号处理, 流水并行化, 重采样, 联合检测估计

Abstract:

Aiming at the high-speed and efficient computing problem of Bernoulli particle filters in embedded application environments, taking the Bernoulli particle filters for joint detection and estimation of radar weak targets as an example, a function modularized and particle size scalable field programmable gate array (FPGA) implementation architecture is proposed. The computing speed of filtering calculation is further improved through approaches as particle state pipelining, layered accumulation and sum, and parallel resampling, etc. Xilinx ZC706 evaluation board on-board testing experiments have demonstrated the good scalability and excellent acceleration ratio of the proposed architecture. When the number of particles is 1 024, the acceleration ratio is about 10⁴ orders of magnitude compared to the Intel Core i3-4130 CPU computing environment. The results have important reference value for the application of Bernoulli particle filtering technology in radar, robotics, and navigation guidance fields.

Key words: Bernoulli particle filter, field programmable gate array (FPGA), real-time signal processing, pipelining parallelization, resample, joint detection and estimation

中图分类号:

TN953

连红飞, 李东升, 蒋彦雯, 范红旗, 肖怀铁, 王国嫣. 一种伯努利粒子滤波器的FPGA实现[J]. 系统工程与电子技术, 2025, 47(2): 398-405.

Hongfei LIAN, Dongsheng LI, Yanwen JIANG, Hongqi FAN, Huaitie XIAO, Guoyan WANG. FPGA implementation of a Bernoulli particle filter[J]. Systems Engineering and Electronics, 2025, 47(2): 398-405.

图/表 13

图1

图2

图3

图4

表1

表2

表3

图5

图6

表4

图7

图8

图9

参考文献 30

1	蔡飞. 雷达弱小目标检测与跟踪技术研究[D]. 长沙: 国防科学技术大学, 2015.
	CAI F. Research on detection and tracking technologies for dim targets in radar[D]. Changsha: National University of Defense Technology, 2015.
2	LIANG T, BATTISTELLI G, CHISCI L, et al. Joint emitter detection and tracking based on the Bernoulli filter[C]//Proc. of the 26th International Conference on Information Fusion, 2023.
3	CHENG X , JI H B , ZHANG Y Q . Effective implementation and improvement of fast labeled multi-Bernoulli filter[J]. Journal of Systems Engineering and Electronics, 2023, 34 (3): 661- 673. doi: 10.23919/JSEE.2023.000030
4	DAVIES E S , GARCIA-FERNANDEZ A F . Information exchange track-before-detect multi-Bernoulli filter for superpositional sensors[J]. IEEE Trans.on Signal Processing, 2024, 72, 607- 621. doi: 10.1109/TSP.2024.3349769
5	ISHTIAQ N , GOSTAR A K , BAB-HADIASHAR A , et al. Interaction-aware labeled multi-Bernoulli filter[J]. IEEE Trans.on Intelligent Transportation Systems, 2023, 24 (11): 11668- 11681. doi: 10.1109/TITS.2023.3294519
6	陈壮壮, 宋骊平. 机动目标跟踪的交互多模型泊松多伯努利混合滤波[J]. 系统工程与电子技术, 2024, 46 (3): 786- 794. doi: 10.12305/j.issn.1001-506X.2024.03.03
	CHEN Z Z , SONG L P . Interacting multiple model Poisson multi-Bernoulli mixture filter for tracking maneuvering targets[J]. Systems Engineering and Electronics, 2024, 46 (3): 786- 794. doi: 10.12305/j.issn.1001-506X.2024.03.03
7	CHISHOLM T , LINS R , GIVIGI S . FPGA-based design for real-time crack detection based on particle filter[J]. IEEE Trans.on Industrial Informatics, 2020, 16 (9): 5703- 5711. doi: 10.1109/TII.2019.2950255
8	KIM D , LEE H , KWON H H , et al. Parallelized particle filter with efficient pipelining on FPGA for real-time ballistic target tracking[J]. IEEE Access, 2023, 11, 104830- 104838. doi: 10.1109/ACCESS.2023.3317896
9	KIM D , HAN Y , LEE H , et al. Accelerated particle filter with GPU for real-time ballistic target tracking[J]. IEEE Access, 2023, 11, 12139- 12149. doi: 10.1109/ACCESS.2023.3238873
10	TANG X Y , FU Z J . CPU-GPU utilization aware energy-efficient scheduling algorithm on heterogeneous computing systems[J]. IEEE Access, 2020, 8, 58948- 58958. doi: 10.1109/ACCESS.2020.2982956
11	SZWOCH G. Performance evaluation of the parallel object tracking algorithm employing the particle filter[C]//Proc. of the IEEE Conference on Signal Processing: Algorithms, Architectures, Arrangements, and Applications, 2016: 119-124.
12	KRISHNA A , SCHAIK A V , THAKUR C S . FPGA implementation of particle filters for robotic source localization[J]. IEEE Access, 2021, 9, 98185- 98203. doi: 10.1109/ACCESS.2021.3094962
13	DIAKITE D , GAC N . X-ray tomography reconstruction acce-lerated on FPGA through high-level synthesis tools[J]. IEEE Trans.on Biomedical Circuits and Systems, 2023, 17 (2): 375- 389. doi: 10.1109/TBCAS.2023.3258879
14	ZHANG J H , HE X , ZHAO G , et al. FPGA implementation for finite-time and fixed-time neurodynamic algorithms in constrained optimization problems[J]. IEEE Trans.on Circuits and Systems Ⅰ: Regular Papers, 2023, 70 (9): 3584- 3597. doi: 10.1109/TCSI.2023.3285933
15	SUN H , DENG Q , LIU X Z , et al. An energy-efficient stream-based FPGA implementation of feature extraction algorithm for LiDAR point clouds with effective local-search[J]. IEEE Trans.on Circuits and Systems Ⅰ: Regular Papers, 2023, 70 (1): 253- 265. doi: 10.1109/TCSI.2022.3212075
16	ZHOU X W , ZHAO D D , GENG Z S , et al. FPGA implementation of non-commensurate fractional-order state-space models[J]. IEEE Trans.on Circuits and Systems Ⅰ: Regular Papers, 2023, 70 (9): 3639- 3652. doi: 10.1109/TCSI.2023.3291368
17	汪敏, 冯一伦, 蒋彦雯, 等. 雷达波形通用调制引擎设计[J]. 系统工程与电子技术, 2023, 45 (6): 1684- 1692. doi: 10.12305/j.issn.1001-506X.2023.06.12
	WANG M , FENG Y L , JIANG Y W , et al. Design of general modulation engine for radar waveform[J]. Systems Engineering and Electronics, 2023, 45 (6): 1684- 1692. doi: 10.12305/j.issn.1001-506X.2023.06.12
18	TAHARA A, HAYASHIDA Y, THU T T, et al. Power performance analysis of FPGA-based particle filtering for realtime object tracking[C]//Proc. of the Conference on Complex, Intelligent, and Software Intensive Systems, 2018: 451-462.
19	HONG S J , CHIN S S , DJURIC P M , et al. Design and implementation of flexible resampling mechanism for high-speed parallel particle filters[J]. Journal of VLSI Signal Processing, 2006, 44 (1/2): 47- 62.
20	EL-HALYM H A A , MAHMOUD I I , HABIB S . Proposed hardware architectures of particle filter for object tracking[J]. EURASIP Journal on Advances in Signal Processing, 2012, 17.
21	MIAO L F , ZHANG J J , CHAKRABARTI C , et al. Efficient Bayesian tracking of multiple sources of neural activity: algorithms and real-time FPGA implementation[J]. IEEE Trans.on Signal Processing, 2013, 61 (3): 633- 647. doi: 10.1109/TSP.2012.2226172
22	ALAM S A , GUSTAFSSON O . Improved particle filter resampling architectures[J]. Journal of Signal Processing Systems, 2020, 92 (6): 555- 568. doi: 10.1007/s11265-019-01489-y
23	AKSHAY A , BOLIC M , SANGJIN H , et al. Generic hardware architectures for sampling and resampling in particle filters[J]. EURASIP Journal on Advances in Signal Processing, 2005 (17): 476167.
24	BOLIC M , DJURIC P M , HONG S J . Resampling algorithms and architectures for distributed particle filters[J]. IEEE Trans.on Signal Processing, 2005, 53 (7): 2442- 2450. doi: 10.1109/TSP.2005.849185
25	SCHWIEGELSHOHN F, OSSOVSKI E, HUBNER M. A fully parallel particle filter architecture for FPGAs[C]//Proc. of the Conference on Applied Reconfigurable Computing, 2015: 91-102.
26	SCHWIEGELSHOHN F , OSSOVSKI E , HV BNER M . A resamp- ling method for parallel particle filter architectures[J]. Microprocessors and Microsystems, 2016, 47, 314- 320. doi: 10.1016/j.micpro.2016.07.017
27	MURRAY L M , LEE A , JACOB P E . Parallel resampling in the particle flter[J]. Journal of Computational and Graphical Statistics, 2016, 25 (3): 789- 805. doi: 10.1080/10618600.2015.1062015
28	辛怀声, 宋鹏汉, 曹晨. 多模型广义标签多伯努利滤波器[J]. 系统工程与电子技术, 2022, 44 (12): 3603- 3613. doi: 10.12305/j.issn.1001-506X.2022.12.03
	XIN H S , SONG P H , CAO C . Multiple model based generali-zed labeled muti-Bernoulli filter[J]. Systems Engineering and Electronics, 2022, 44 (12): 3603- 3613. doi: 10.12305/j.issn.1001-506X.2022.12.03
29	吴孙勇, 邹宝红, 薛秋条, 等. 基于单快拍空间平滑的多伯努利DOA跟踪算法[J]. 系统工程与电子技术, 2021, 43 (9): 2430- 2438.
	WU S Y , ZOU B H , XUE Q T , et al. DOA tracking algorithm based on single snapshot spatial smoothing with multi-Bernoulli[J]. Systems Engineering and Electronics, 2021, 43 (9): 2430- 2438.
30	GNING A , RISTIC B , MIHAYLOVA L . Bernoulli particle/box-particle filters for detection and tracking in the presence of triple measurement uncertainty[J]. IEEE Trans.on Signal Processing, 2012, 60 (5): 2138- 2151. doi: 10.1109/TSP.2012.2184538

参数名称	取值
存活粒子个数N	512
新生粒子个数B	512
输入序列长度L/点	250
参数名称	取值
采样时间/ms	5
输入信号信噪比/dB	4
系统时钟频率/MHz	200

计算单元名称	时间
更新单元	45.30
重采样单元	24.96
状态提取单元	21.71
随机数产生单元	5.16
存活粒子预测单元	4.54
粒子新生单元	5.54

资源类型	消耗量/个	占比/%
FF	28 598	6.5
LUT	45 632	20.8
BRAM	84	7.7
DSP Slices	132	14.6

计算环境	均值	误差
FPGA	-500.81	0.81
CPU	-501.25	1.25

[1]	陈昊然, 王天昊, 路美娜, 宋茂新, 罗环, 吴晓宇, 骆冬根, 裘桢炜. 基于HLS的高精度位移测量算法的硬件加速设计[J]. 系统工程与电子技术, 2025, 47(2): 341-351.
[2]	刘永庆, 刘鹏, 云日升, 张祥坤, 王特. 海洋4A散射计幅相校正算法FPGA优化实现[J]. 系统工程与电子技术, 2024, 46(8): 2554-2562.
[3]	汪敏, 冯一伦, 蒋彦雯, 范红旗. 雷达波形通用调制引擎设计[J]. 系统工程与电子技术, 2023, 45(6): 1684-1692.
[4]	李睿峰, 许爱强, 孙伟超, 吴阳勇. 基于样本重采样的电路非平衡数据预处理方法[J]. 系统工程与电子技术, 2020, 42(11): 2654-2660.
[5]	张一迪, 王培志, 陆起涌, 张建秋. 异常数据恒虚警检测的非参数方法[J]. 系统工程与电子技术, 2019, 41(5): 964-971.
[6]	王进花, 曹洁, 李伟. 智能优化的代价评估粒子滤波算法[J]. 系统工程与电子技术, 2017, 39(12): 2857-2862.
[7]	安辉耀1, 刘敦伟1,2, 耿瑞华3, 曾和平4, 赵林欣1. #br# 量子通信系统中基于FPGA的偏振控制[J]. 系统工程与电子技术, 2016, 38(8): 1917-1921.
[8]	王钊, 顾红, 苏卫民, 樊劲宇, 陈金立. 宽带高斯谱噪声雷达高速目标检测[J]. 系统工程与电子技术, 2015, 37(6): 1266-1272.
[9]	樊养余，于泽琦，袁永金，吕国云. 基于FPGA的高性能D类功放控制器#br# 设计与实现[J]. 系统工程与电子技术, 2014, 36(4): 630-636.
[10]	唐林波, 陶芬芳, 赵保军, 刘嘉骏. 基于FPGA的实时整数霍夫变换[J]. Journal of Systems Engineering and Electronics, 2012, 34(3): 610-613.
[11]	刘光祖，王建新，徐达龙. 数字信道化接收机高效结构的设计与实现[J]. Journal of Systems Engineering and Electronics, 2012, 34(2): 391-395.
[12]	罗海坤，吴嗣亮，王永庆. 基于改进ziggurat算法的高斯白噪声发生器[J]. Journal of Systems Engineering and Electronics, 2011, 33(4): 879-883.
[13]	黄杰文，祁海明，李早社，禹卫东. 星载SAR中频数字接收机的FPGA设计与实现[J]. Journal of Systems Engineering and Electronics, 2011, 33(4): 769-773.
[14]	段军棋，何子述. 多频雷达信号高分辨多普勒处理[J]. Journal of Systems Engineering and Electronics, 2010, 32(3): 475-477.
[15]	杨保平，陈永光，陈军，徐忠富. 基于FPGA的战术数据链中继传输复接技术[J]. Journal of Systems Engineering and Electronics, 2010, 32(12): 2628-2631.

一种伯努利粒子滤波器的FPGA实现

FPGA implementation of a Bernoulli particle filter

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 30

相关文章 15

编辑推荐

Metrics

本文评价