南京大学学报(自然科学版) ›› 2021, Vol. 57 ›› Issue (5): 838–846.doi: 10.13232/j.cnki.jnju.2021.05.014

• • 上一篇    

基于多站点、多时间注意力机制的电磁强度时空关联分析与可视化

徐樱笑, 资文杰, 宋洁琼, 邵瑞喆, 陈浩()   

  1. 国防科技大学电子科学学院，长沙，410073
  • 收稿日期:2021-06-26 出版日期:2021-09-29 发布日期:2021-09-29
  • 通讯作者: 陈浩 E-mail:hchen@nudt.edu.cn
  • 作者简介:E⁃mail：hchen@nudt.edu.cn
  • 基金资助:
    国家自然科学基金(61806211);湖南省自然科学基金(2020JJ4103)

Spatio⁃temporal correlation analysis and visualization of electromagnetic intensity based on multi⁃site and multi⁃ temporal attention mechanism

Yingxiao Xu, Wenjie Zi, Jieqiong Song, Ruizhe Shao, Hao Chen()   

  1. College of Electronic Science and Technology，National University of Defense Technology，Changsha，410073，China
  • Received:2021-06-26 Online:2021-09-29 Published:2021-09-29
  • Contact: Hao Chen E-mail:hchen@nudt.edu.cn

RichHTML

21

PDF (PC)

814

摘要：

从监测端挖掘多监测站点间电磁信息隐含的时空关联对掌握电磁频谱资源分布和使用状况、支撑电磁频谱资源精细化管理具有重要意义.目前的方法孤立地对各监测站点采集的海量数据进行预测分析，没有有效挖掘多个站点间的电磁信息变化规律，没有考虑人类活动规律的周期性，更难以利用站点间的时域、空域关联给出多站点的电磁强度信息联合预测，不能结合GIS系统给出直观、实时可视化的频谱态势地图.针对以上问题，将多个监测站点建模为图结构，提出一种基于多站点、多时间注意力机制的图神经网络MMGCN （Multi?site and Multi?temporal Attention?based Graph Convolutional Neural Network），挖掘多站点电磁强度间隐含的时空关联，提升联合预测的准确性；基于多站点联合预测数据，设计一种并行克里金插值算法，对其他未布设监测点处的电磁强度信息进行空间插值生成电磁频谱态势地图，为电磁频谱精细化管理提供直观的态势预判与决策依据.

关键词: 多监测站点, 多时间注意力机制, 图神经网络, 电磁强度预测, 电磁频谱态势地图

Abstract:

Mining the spatial?temporal correlation of electromagnetic information from multiple monitor sites is of great significance for mastering the distribution and usage of electromagnetic spectrum resources. In traditional approaches，the prediction of electromagnetic intensity information mainly uses forecasting methods to predict and analyze the massive monitoring data collected by sites in isolation without correlation analysis in the time?domain or the spatial?domain. The prediction ignores to mine the law among multiple stations to give the joint prediction of electromagnetic intensity information and consider the periodicity of human activity law. Therefore，it is difficult to provide a real?time spectrum situation map based on the forecast data and GIS for managing electromagnetic spectrum resources. To address the challenges mentioned above，we firstly model the monitoring information by multi?site as a graph and propose a graph neural network based on multi?site and multi?temporal attention mechanism (MMGCN). MMGCN mines the implied spatio?temporal correlation between multi?site to improve joint prediction accuracy. Secondly，a parallel Kriging interpolation algorithm is designed to spatially interpolate the electromagnetic intensity information at other undeployed monitoring sites to generate a successive spectrum map，which provides intuitive situation prediction and decision?making basis for further management of electromagnetic spectrum.

Key words: multiple monitoring sites, multi?temporal attention?based mechanism, Graph Neural Network (GNN), electromagnetic intensity prediction, spectral situation map

中图分类号: 

  • TP301

图1

监测站点分布示意图"

图2

MMGCN图神经网络结构示意图"

图3

克里金?双线性插值方法示意图"

表1

模型预测结果的指标统计"

模型MAERMSE
HA2.6723.171
LASSO2.3212.879
LSTM0.3740.513
GCNN1.7912.355
ASTGNN0.2160.082
本文MMGCN0.0990.014

表2

算法运行时间对比"

模型

时间(s)

时间复杂度

串行克里金插值算法

283.66

Omn2

并行克里金插值算法

274.47

Omn2

并行克里金?双线性插值算法

11.90

Om+n2

图4

海口市及周边地区电磁频谱态势图"

1 吴启晖，任敬. 电磁频谱空间认知新范式：频谱态势. 南京航空航天大学学报，2016，48(5)：625-632.
Wu Q H，Ren J. New Paradigm of electromagnetic spectrum space：Spectrum situation. Journal of Nanjing University of Aeronautics & Astronautics，2016，48(5)：625-632.
2 Yin L，Yin S X，Hong W J，et al. Spectrum behavior learning in cognitive radio based on artificial neural network∥2011 Military Communications Conference. Baltimore，MD，USA：IEEE，2011：25-30.
3 Li H Y，Xu X D，Wu B X，et al. Multilayer feedforward neural network based efficient spectrum occupancy prediction scheme for Cognitive Radio system. Journal of Computational Information Systems，2014，10(10)：4017-4028.
4 刘栩楠. 城市无线电监测数据挖掘方法及应用. 硕士学位论文. 成都：电子科技大学，2020.
Liu Xunan. Urban radio monitoring data mining method and its application. Master Dissertation. Chengdu：University of Electronic Science and Technology，2020.
5 韩仿仿，杨娟. 短波监测工作中的信号识别与分析. 中国无线电，2015(4)：44-46.
Han F F，Yang J. Signal identification and analysis in short?wave monitoring. China Radio，2015(4)：44-46.
6 Cotton M，Wepman J，Kub J，et al. An overview of the NTIA/NIST spectrum monitoring pilot program∥Proceedings of 2015 IEEE Wireless Communications and Networking Conference Workshops. Los Alamitos，LA，USA：IEEE，2015：217-222.
7 郭仕剑，李坡，胡昆明等. 实时频谱分析技术在雷达信号侦察中的应用. 信号处理，2011，27(8)：1229-1234.
Guo S J，Li P，Hu K M，et al. Application of real?time spectrum analysis technology in radar reconnaissance. Signal Processing，2011，27(8)：1229-1234.
8 夏海洋，查淞，黄纪军等. 电磁频谱地图构建方法研究综述及展望. 电波科学学报，2020，35(4)：445-456.
Xia H Y，Zha S，Huang J J，et al. Survey on the construction methods of spectrum map. Chinese Journal of Radio Science，2020，35(4)：445-456.
9 杜浩，毕大平. 复杂电磁空间中电磁频谱管理的应对策略分析. 电子信息对抗技术，2014，29(5)：10-13. （Du H，Bi D P. The analysis of electromagnetic spectrum management strategy in complex electromagnetic environment. Electronic Information
Warfare Technology，2014，29(5)：10-13.
10 张海鸥，潘超，夏远芬等. 移动通信基站电磁辐射时空分布及衰减特征. 电力环境保护，2009，25(4)：55-57. （Zhang H O，Pan C，Xia Y F，et al. Temporal and spatial distribution and attenuation characteristics of electromagnetic radiation in mobile communication base station. Electric Power Environment
Protection，2009，25(4)：55-57.
11 刘栩楠，石荣，何彬彬等. 基于分段马尔可夫模型的频谱占用预测方法. 无线电工程，2020，50(8)：690-695. （Liu X N，Shi R，He B B，et al. Spectrum occupancy prediction method based on segmented
Markov model. Radio Engineering，2020，50(8)：690-695.
12 Yin S X，Chen D W，Zhang Q，et al. Prediction?based throughput optimization for dynamic spectrum access. IEEE Transactions on Vehicular Technology，2011，60(3)：1284-1289.
13 Wang Z，Salous S. Spectrum occupancy statistics and time series models for cognitive radio. Journal of Signal Processing Systems，2011，62(2)：145-155.
14 Wu Z H，Pan S R，Chen F W，et al. A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems，2021，32(1)：4-24.
15 Bruna J，Zaremba W，Szlam A，et al. Spectral networks and locally connected networks on graphs. 2013，arXiv:.
16 Battaglia P W，Hamrick J B，Bapst V，et al. Relational inductive biases，deep learning，and graph networks. 2018，arXiv：.
17 Zhang C H，Yu J J Q，Liu Y. Spatial?temporal graph attention networks：A deep learning approach for traffic forecasting. IEEE Access，2019(7)：166246-166256.
18 Jayawickrama B A，Dutkiewicz E，Oppermann I，et al. Iteratively reweighted compressive sensing based algorithm for spectrum cartography in cognitive radio networks∥2014 IEEE Wireless Communications and Networking Conference. Istanbul，Turkey：IEEE，2014：719-724.
19 Tang M Y，Ding G R，Wu Q H，et al. A joint tensor completion and prediction scheme for multi?dimensional spectrum map construction. IEEE Access，2016(4)：8044-8052.
20 Scholl P M，Kohlbrecher S，Sachidananda V，et al. Fast indoor radio?map building for RSSI?based localization systems∥2012 9th International Conference on Networked Sensing. Antwerp，Belgium：IEEE，2012：1-2.
21 Lee M，Han D. Voronoi tessellation based interpolation method for Wi?Fi radio map construction. IEEE Communications Letters，2012，16(3)：404-407.
22 Romero D，Kim S J，Giannakis G B. Stochastic semiparametric regression for spectrum cartography∥2015 IEEE 6th International Workshop on Computational Advances in Multi?Sensor Adaptive Processing. Cancun，Mexico：IEEE，2015：513-516.
23 Tang M Y，Zheng Z，Ding G R，et al. Efficient TV white space database construction via spectrum sensing and spatial inference∥2015 IEEE 34th International Performance Computing and Communications Conference. Nanjing，China：IEEE，2015：1-5.
24 Geng X P，Hu P. Spatial interpolation method of scalar data based on raster distance transformation of map algebra∥International Symposium on Intelligence Information Processing and Trusted Computing. Huanggang，China：IEEE，2010：188-191.
25 Ding G R，Wang J L，Wu Q H，et al. Joint spectral?temporal spectrum prediction from incomplete historical observations∥2014 IEEE Global Conference on Signal and Information Processing. Atlanta，GA，USA：IEEE，2015：1325-1329.
26 Gandy S，Recht B，Yamada I. Tensor completion and low?n?rank tensor recovery via convex optimization. Inverse Problems，2011，27(2)：025010.
27 Mao D L，Shao W，Qian Z P，et al. Constructing accurate radio environment maps with Kriging interpolation in cognitive radio networks∥2018 Cross Strait Quad?Regional Radio Science and Wireless Technology Conference. Xuzhou，China：IEEE，2018：1-3.
28 Yu B，Yin H T，Zhu Z X. Spatio?temporal graph convolutional networks：A deep learning framework for traffic forecasting∥The 27th International Joint
Conference on Artificial Intelligence. Stockholm，
Sweden：IJCAI.org，2018：3634-3640.
29 Zhao L，Song Y J，Zhang C，et al. T?gcn：A temporal graph convolutional network for traffic prediction. IEEE Transactions on Intelligent Transportation Systems，2019，21(9)：3848-3858.
30 Zhou J，Cui G Q，Hu S D，et al. Graph neural networks：A review of methods and applications. 2018，arXiv:.
31 Henaff M，Bruna J，LeCun Y. Deep convolutional networks on graph?structured data. 2015，arXiv:.
32 Qian J，Comin M，Pianura L. Data?driven smart bike?sharing system by implementing machine learning algorithms∥2018 6th International Conference on Enterprise Systems. Limassol，Cyprus：IEEE，2018：50-55.
33 Hochreiter S，Schmidhuber J. Long short?term memory. Neural Computation，1997，9(8)：1735-1780.
34 Lin L，He Z B，Peeta S. Predicting station?level hourly demand in a large?scale bike?sharing network：A graph convolutional neural network approach. Transportation Research Part C：Emerging Technologies，2018(97)：258-276.
35 Zhou H，Ren D C，Xia H X，et al. AST?GNN：An attention?based spatio?temporal graph neural network for Interaction?aware pedestrian trajectory prediction. Neurocomputing，2021(445)：298-308.
[1] 郝昱猛, 马文明, 王冰. 基于特定用户约束的概率矩阵分解算法[J]. 南京大学学报(自然科学版), 2021, 57(5): 818-827.
[2] 杨静, 赵文仓, 徐越, 冯旸赫, 黄金才. 一种基于少样本数据的在线主动学习与分类方法[J]. 南京大学学报(自然科学版), 2021, 57(5): 757-766.
[3] 陈炳鑫, 陈黎飞. 符号序列的预训练HMM分类方法[J]. 南京大学学报(自然科学版), 2021, 57(1): 52-58.
[4] 李亚重,杨有龙,仇海全. 一种基于嵌入式的弱标记分类算法[J]. 南京大学学报(自然科学版), 2020, 56(4): 549-560.
[5] 王丽娟,丁世飞,丁玲. 基于迁移学习的软子空间聚类算法[J]. 南京大学学报(自然科学版), 2020, 56(4): 515-523.
[6] 李俊余, 李星璇, 王霞, 吴伟志. 基于三元因子分析的三元概念约简[J]. 南京大学学报(自然科学版), 2020, 56(4): 480-493.
[7] 何昕,徐珺平,赵登吉. 基于社交网络的分布式机制设计[J]. 南京大学学报(自然科学版), 2020, 56(1): 107-115.
[8] 袁燕,陈伯伦,朱国畅,花勇,于永涛. 基于社区划分的空气质量指数（AQI）预测算法[J]. 南京大学学报(自然科学版), 2020, 56(1): 142-150.
[9] 刘亮,何庆. 基于改进蝗虫优化算法的特征选择方法[J]. 南京大学学报(自然科学版), 2020, 56(1): 41-50.
[10] 李子龙,周勇,鲍蓉. AdaBoost图像到类距离学习的图像分类方法[J]. 南京大学学报(自然科学版), 2020, 56(1): 51-56.
[11] 袁友宏,刘欣,鲍蕾. 求解非凸截断L 1⁃SVM的多阶段非精确线搜割平面方法[J]. 南京大学学报(自然科学版), 2020, 56(1): 98-106.
[12] 王霞, 谭斯文, 李俊余, 吴伟志. 基于条件属性蕴含的概念格构造及简化[J]. 南京大学学报(自然科学版), 2019, 55(4): 553-563.
[13] 王博闻, 史江峰, 史逝远, 张伟杰, 马晓琦, 赵业思. 基于遥感数据定位老龄树群[J]. 南京大学学报(自然科学版), 2019, 55(4): 699-707.
[14] 陈晓冬, 底晓强, 李锦青. 基于多混沌和分数Fourier的光学图像加密算法[J]. 南京大学学报(自然科学版), 2019, 55(2): 251-263.
[15] 洪思思，曹辰捷，王　喆*，李冬冬. 基于矩阵的AdaBoost多视角学习[J]. 南京大学学报(自然科学版), 2018, 54(6): 1152-1160.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!

版权所有 © 2021 《南京大学学报(自然科学)》

地址：江苏省南京市鼓楼区汉口路22号，南京大学学报编辑部，210093

电话：025-83592704 , 传真：025-83592704        技术支持：北京玛格泰克科技发展有限公司