利用 DPRNN 实现交织重数未知的 ADS-B 信号分离

taptap下载安装安卓学报 ›› 2025, Vol. 43 ›› Issue (3): 75-80.

利用 DPRNN 实现交织重数未知的 ADS-B 信号分离

taptap下载安装安卓电子信息与自动化学院，天津 300300

收稿日期:2023-01-17 修回日期:2023-05-09 出版日期:2025-07-12 发布日期:2025-07-12
作者简介:王文益（1980— ），男，湖北襄阳人，教授，博士，研究方向为卫星导航抗干扰、自适应信号处理
基金资助:
国家自然科学基金项目（U2133204）；天津市研究生科研创新项目（2021YJS023）

Separating ADS-B signal with unknown interweaving multiplicity

using DPRNN

College of Electronical Information and Automation, CAUC, Tianjin 300300, China

Received:2023-01-17 Revised:2023-05-09 Online:2025-07-12 Published:2025-07-12

摘要/Abstract

摘要：

多条信号之间的交织是广播式自动相关监视（ADS-B，automatic dependent surveillance-broadcast）系统中不

可避免的问题。已有的单天线 ADS-B 信号解交织方法，大多需要精确的估计交织重数、相对时延等参数，相

应的估计精度往往严重影响最终的分离性能。为解决该问题，本文利用双路径递归神经网络（DPRNN，dualpath recurrent neural network）在无须估计任何参数的前提下对 ADS-B 接收信号进行分离。网络输入为交织

重数未知的 ADS-B 接收信号，包括纯噪声信号、无交织信号、两重交织信号和三重交织信号。网络固定输出

3 条信号，当交织信号中包含 i 条 ADS-B 信号时，对应的源信号包含 i 条 ADS-B 信号和（3 - i）条噪声信

号。仿真实验证明，当信号的交织重数不大于三重时，分离正确率达到 95% 以上。这一研究为 ADS-B 信号

交织分离提供了一种新的解决方案，具有一定的实用价值。

关键词:

广播式自动相关监视（ADS-B）, 信号交织, 信号分离, 双路径递归神经网络

Abstract:

The interweaving of multiple signals is an inevitable problem in automatic dependent surveillance -broadcast

(ADS-B) system. Existing single-antenna ADS-B signal de-interweaving methods mostly require accurate estimation of parameters such as the interweaving multiplicity and relative delay, and the estimation accuracy often seriously affects the final separation performance. To address this problem, this paper uses dual-path recurrent neural

network (DPRNN) to separate ADS-B reception signals without the need to estimate any parameters. The network input is ADS-B signals with unknown interweaving multiplicity, including pure noise signal, no interweaving

signal, double interweaving signal and triple interweaving signal. The network is fixed to output three signals, and

when the interweaving signal contains i ADS-B signals, the corresponding source signals have i ADS-B signals

and (3 - i) noise signals. Simulation experiments show that the separation accuracy is above 95% when the interweaving multiplicity is no more than three. This study provides a new solution for ADS-B signal interweaving separation, which has practical value.

Key words:

automatic dependent surveillance-broadcast (ADS-B), signal interweaving, signal separation, dual-path recurrent neural network (DPRNN)

中图分类号:

王文益, 王建设.

利用 DPRNN 实现交织重数未知的 ADS-B 信号分离

[J]. taptap下载安装安卓学报, 2025, 43(3): 75-80.

WANG Wenyi, WANG Jianshe.

Separating ADS-B signal with unknown interweaving multiplicity

using DPRNN

[J]. Journal of Civil Aviation University of China, 2025, 43(3): 75-80.

参考文献 16

	[1]
	LEONARDI M, PIRACCI E, GALATI G. ADS-B jamming mitigation: a
	solution based on a multichannel receiver[J]. IEEE Aerospace and Electronic Systems Magazine, 2017, 32(11): 44-51.
	[2]
	RTCA. Minimum operational performance standards for 1090 MHz ex－
	tended squitter ADS-B: RTCA DO-260B[S]. Washington DC: RTCA,
	2009.
	[3]
	VAN DER PRYT R, VINCENT R. A simulation of signal collisions over
	the North Atlantic for a spaceborne ADS-B receiver using Aloha proto－
	col[J]. Positioning, 2015, 6(3): 23-31.
	[4]
	WANG W Y, WU R B, LIANG J L． ADS-B signal separation based on
	blind adaptive beamforming[J]. IEEE Transactions on Vehicular Technology, 2019, 68(7): 6547-6556.
[5]	吴仁彪, 吴琛琛, 王文益. 基于累加分类的 ADS-B 交织信号处理方
法[J]	信号处理, 2017, 33(4): 572-576.
	[6]
	YU S Q, CHEN L H, LI S T, et al. Separation of space-based ADS-B
	signals with single channel for small satellite[C]//2018 IEEE 3rd International Conference on Signal and Image Processing (ICSIP), July 13-15,
	2018, Shenzhen, China． IEEE, 2018: 315-321.
	[7]
	LI K D, KANG J, REN H, et al． A reliable separation algorithm of ADSB signal based on time domain[J]. IEEE Access, 2021, 9: 88019-88026.
	[8]
	GALATI G, PETROCHILOS N, PIRACCI E G. Degarbling Mode S replies received in single channel stations with a digital incremental im－
	provement[J]. IET Radar, Sonar & Navigation, 2015, 9(6): 681-691.
[9]	王文益, 邵宇识. 基于改进单天线投影算法的广播式自动相关监视
	信号分离[J]. 电子与信息学报, 2020, 42(11): 2720-2726.
[10]	BI Y, LI C K． Multi-scale convolutional network for space-based ADSB signal separation with single antenna[J]. Applied Sciences, 2022, 12
(17)	: 8816.
[11]	YANG Y, ZHANG H P, ZHA H R, et al． ADS-B signal separation via
	complex neural network[C]//International Conference on Mobile Multi－
	media Communications. Cham: Springer International Publishing, 2021:
	679-688．
[12]	WANG W Y, LIU J, LIANG J L． Single antenna ADS-B overlapping
	signals separation based on deep learning[J]． Digital Signal Processing,
	2023, 132: 103804.
[13]	SHI X J, CHEN Z R, WANG H, et al. Convolutional LSTM network: a
	machine learning approach for precipitation nowcasting[J]. Advances in
	Neural Information Processing Systems, 2015: 802-810.
[14]	LUO Y, CHEN Z, YOSHIOKA T． Dual-path RNN: efficient long se－
	quence modeling for time-domain single channel speech separation[C]//
	ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech
	and Signal Processing (ICASSP), May 4-8, 2020． Barcelona, Spain． IEEE,
	2020: 46-50.
[15]	LOPEZ PINAYA W H, VIEIRA S, GARCIA-DIAS R, et al. Autoen－
	coders[M]//Machine Learning. Ansterdam: Elsevier, 2020: 193-208.
[16]	SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for
	large -scale image recognition [C]//3rd International Conference on
	Learning Representations (ICLR), May 7-9, 2015, San Diego, CA, USA.
	2015: 1-14.