空中交通管理系统网络安全研究综述与展望

王布宏; 罗鹏; 阳勇; 赵正阳; 董若辰; 关永健

doi:10.11999/JEIT240966

空中交通管理系统网络安全研究综述与展望

doi: 10.11999/JEIT240966

空军工程大学信息与导航学院西安 710077

基金项目: 国家自然科学基金(62472437)

详细信息

作者简介:
王布宏：男，教授，研究方向为空管信息系统安全和渗透测试

罗鹏：男，博士生，研究方向为空管网络异常检测和人工智能安全

阳勇：男，博士生，研究方向为云计算和网络功能虚拟化

赵正阳：男，博士生，研究方向为模式识别和人工智能安全

董若辰：男，博士生，研究方向为空管网络和信息物理系统安全

关永健：男，硕士，研究方向为网络安全与人工智能

通讯作者:
王布宏　wbhgroup@aliyun.com

中图分类号: TN915.08
计量
- 文章访问数: 346
- HTML全文浏览量: 105
- PDF下载量: 70
- 被引次数: 1
出版历程
- 收稿日期: 2024-10-29
- 修回日期: 2025-02-20
- 网络出版日期: 2025-03-08

A Review and Prospect of Cybersecurity Research on Air Traffic Management Systems

Information and Navigation School, Air Force Engineering University, Xi’an 710077, China

Funds: The National Natural Science Foundation of China (62472437)

摘要

摘要: 空中交通管理系统是关乎空天安全和人民生命财产安全的国家大型关键信息基础设施。随着信息化、网络化和智能化技术的广泛应用，现代空中交通管理系统已经演化成为由多利益相关方异构融合的空天地海一体化网络。尽管系统的开放性和连接性提升了空中交通管理效率，但也引入了新的网络安全威胁，扩大了系统攻击面，使得网络安全生态复杂且形势严峻。该文以资产梳理、威胁分析、攻击建模、防御机制为主线，从不同的利益相关方(Stakeholders)出发，如电子使能飞机、空中交通管理(CNS/ATM)、智慧机场、智能决策等方面对空中交通管理系统的网络安全研究现状进行了全面系统的综述，并提出了动态网络安全分析、攻击影响传播建模、人在回路网络安全分析、分布式入侵检测系统等方面的研究问题与挑战。
- 空中交通管理系统 /
- 网络安全 /
- 研究综述 /
- 研究挑战
Abstract: Significance The air traffic management system is a critical national infrastructure that impacts both aerospace security and the safety of lives and property. With the widespread adoption of information, networking, and intelligent technologies, the modern air traffic management system has evolved into a space-air-ground-sea integrated network, incorporating heterogeneous systems and multiple stakeholders. The network security of the system can no longer be effectively ensured by device redundancy, physical isolation, security by obscurity, or human-in-the-loop strategies. Due to the stringent requirements for aviation airworthiness certification, the implementation of new cybersecurity technologies is often delayed. New types of cyberattacks, such as advanced persistent threats and supply chain attacks, are increasingly prevalent. Vulnerabilities in both hardware and software, particularly in embedded systems and industrial control systems, are continually being exposed, widening the attack surface and increasing the number of potential attack vectors. Cyberattack incidents are frequent, and the network security situation remains critical. Progress The United States’ Next Generation Air Transportation System (NextGen), the European Commission’s Single European Sky Air Traffic Management Research (SESAR), and the Civil Aviation Administration of China have prioritized cybersecurity in their development plans for next-generation air transportation systems. Several countries and organizations, including the United States, Japan, China, the European Union, and Germany, have established frameworks for the information security of air traffic management systems. Although network and information security for air traffic management systems is gaining attention, many countries prioritize operational safety over cybersecurity concerns. Existing security specifications and industry standards are limited in addressing network and information security. Most of them focus on top-level design and strategic directions, with insufficient attention to fundamental theories, core technologies, and key methodologies. Current review literature lacks a comprehensive assessment of assets within air traffic management systems, often focusing only on specific components such as aircraft or airports. Furthermore, research on aviation information security mainly addresses traditional concerns, without fully considering the intelligent and dynamic security challenges facing next-generation air transportation systems. Conclusions This paper comprehensively examines the complexity of the cybersecurity ecosystem in air traffic management systems, considering various entities such as electronic-enabled aircraft, communication, navigation, Surveillance/Air Traffic Management (CNS/ATM), smart airports, and intelligent computing. It focuses on asset categorization, information flow, threat analysis, attack modeling, and defense mechanisms, integrating dynamic flight phases to systematically review the current state of cybersecurity in air traffic management systems. Several scientific issues are identified that must be addressed in constructing a secure ecological framework for air traffic management. Based on the Adversarial Tactics, Techniques, and Common Knowledge (ATT&CK) model, this paper analyzes typical attack examples related to the four ecological entities (Figs. 7, 9, 12, and 14) and constructs an ATT&CK matrix for air traffic management systems (Fig. 15). Additionally, with the intelligent development goal of next-generation air transportation systems as a guide, ten typical applications of intelligent air traffic management are outlined (Fig. 13, Table 11), with a systematic analysis of the attack patterns and defense mechanisms of their intelligent algorithms (Tables 12, 13). These findings provide theoretical references for the development of smart civil aviation and the assurance of cybersecurity in China. Prospects Currently, the cybersecurity ecosystem of air traffic management systems is highly complex, with unclear mechanisms, indistinct boundaries for cybersecurity assets, and incomplete security assurance requirements. Moreover, there is a lack of comprehensive, systematic, and holistic cybersecurity design and defense mechanisms, which limits the ability to counter various subjective, human-driven, and emerging types of malicious cyberattacks. This paper highlights key research challenges in areas such as dynamic cybersecurity analysis, attack impact propagation modeling, human-in-the-loop cybersecurity analysis, and distributed intrusion detection systems. Cybersecurity analysis of air traffic management systems should be conducted within the dynamic operational environment of a space-air-ground-sea integrated network, accounting for the cybersecurity ecosystem and analyzing it across different spatial and temporal dimensions. As aircraft are cyber-physical systems, cybersecurity threat analysis should focus on the interrelated propagation mechanisms between security and safety, as well as their cascading failure models. Furthermore, humans serve as the last line of defense in cybersecurity. When performing threat modeling and risk assessment for avionics systems, it is crucial to fully incorporate “human-in-the-loop” characteristics to derive comprehensive and objective conclusions. Finally, the design, testing, certification, and updating of civil aviation avionics systems are constrained by strict airworthiness requirements, preventing the rapid implementation of advanced cybersecurity technologies. Distributed anomaly detection systems, however, currently represent an effective technical approach for combating cyberattacks in air traffic management systems.
- Air traffic management system /
- Cyber security /
- Research review /
- Research challenges

HTML全文

1. 引言

图像显著性检测的目的是模拟人的视觉特点，提取图像中人类更加关注的区域^[1]。进行全景图像显著性检测，不仅可以提高全景图像的压缩效率，减少传输带宽，而且对全景图像编辑起着至关重要的作用。此外还有力支撑了图像分割、图像检索、目标追踪识别和机器人导航^[2]等计算机视觉任务。

全景图像的显著性检测中，对图像特征的提取效果直接影响最终的显著性检测效果。Zhang等人^[3]提出了一种球形卷积，并验证了球形U型网络对全景视频显著性检测的有效性。Coors等人^[4]提出了一种可以避免极点过采样的球形卷积。Martin等人^[5]提出的显著性检测的方法，主观检测结果好，收敛较缓慢。

全景图像显著性检测时，有多种投影方式，如等矩形投影和立方体投影等。Dai等人^[6]将全景图像以立方体的格式输入扩展卷积神经网络，同时处理立方体的6个面进行全景图像的显著性检测。Monroy等人^[7]提出了对卷积神经网络(Convolutional Neural Network, CNN)的架构扩展，以端到端的方式对全景图像进行显著性检测。以上方法将全景图像投影成2D图像的过程会造成不同程度的信息缺失，因此最终显著性检测结果的性能指标也会受其影响。

Dahou等人^[8]提出用于全景视频显著性检测的基于新注意力的显著性模型(a novel ATtention based Saliency model, ATSal)，该模型对全局视觉静态注意力进行显式编码。Zhu等人^[9]针对显著性检测过程中，全景图像数据集规模小的问题，提出注意力感知特征融合网络。上述方法凭借注意力机制提高了全景图像显著性检测结果的性能指标得分，但这些指标还有上升的空间。

为了提高检测结果的精度，Chao等人^[10]提出从全景图像的3个不同视场(Field of Views, FoV)的每个视口提取特征再融合的显著性检测方法。该方法在性能指标上能取得较好的结果，但是具有很高的计算量。

综上所述，以上方法都有各自的优势，但存在模型收敛速度慢、实际使用受限、全景图像投影成2D图像造成失真以及高计算量等问题。为了解决上述问题，本文提出一种新的全景图像显著性检测网络：基于鲁棒视觉变换和多注意力的U型网络(U-Net with Robust vision transformer and Multiple attention modules, URMNet)。主要贡献如下：

(1) 提出URMNet网络模型，与目前主流全景图像显著性检测模型相比，进一步提升了全景图像显著性检测评价指标。

(2) 提出鲁棒视觉变换模块 (Robust Vision Transformer, RVT )，采用卷积嵌入的方式，通过调整特征图的空间和通道维度，降低了分辨率，解决了网络准确度饱和与性能退化的问题，增强了模型的鲁棒性；提出多注意力模块 (Multiple Attention, MA)，通过融合多维度注意力，提升网络的特征提取能力，提高显著性检测精确度。

(3) 提出更简洁的纬度加权损失函数Loss，加快模型收敛速度，提升全景图像显著性检测效果。

(4) 对原始全景图像显著性检测数据集，即全景图像注意(Attention on Omnidirectional Images, AOI)^[11]、Salient360^[12]分别进行数据增强得到新的数据集，即增强全景图像注意(Augment AOI, AAOI)、增强显著360(Augment Salient360, ASalient360)，大幅增多数据集图像数量。本文模型在两种类型的数据集上达到了预期的效果，证明了模型的有效性和泛化能力。

2. 本文模型

2.1 URMNet网络概述

本文提出一种基于鲁棒视觉变换(RVT)和多注意力(MA)的全景图像显著性检测网络URMNet，如图1所示。URMNet是类U型结构，由编码器、解码器、RVT模块和MA模块组成。URMNet的工作流程如下：输入全景图像到编码模块，首先用球形卷积(sphere convolution)^[4]进行特征提取，然后进行批归一化(Batch Normalization, BN)和修正线性单元(Rectified Linear Unit, ReLU)激活操作，得到第1尺度特征图，再进行球形池化。如此循环4次可得到5种尺度的特征图，这些特征图包含浅层的细节信息和深层的语义信息。将前4种尺度的特征图送入由4个rvt子模块组成的RVT模块，RVT模块可以提取4种尺度特征图所包含的显著信息。同时，最小尺度的特征图通过MA模块的多注意力机制有选择地融合空间和通道显著信息。融合后的显著信息经过上采样后送入解码器首先与RVT模块对应尺度的输出进行拼接，然后经过球形卷积、批归一化BN以及ReLU激活操作进一步细化生成的聚合特征，按此过程一共重复4次逐渐生成精确的显著图。

图 1 URMNet示意图

下载: 全尺寸图片幻灯片

2.2 rvt子模块

为了提高特征的提取速度并兼顾全景图像的全局特征，本文采用包含了多通道自注意力机制的变换器(图1中变换器1)。经过实验，随着RVT-Block(见图2)块数的增多，模型的检测精度会提高，同时在变换器后期降低空间分辨率有利于提高模型的鲁棒性^[13]。因此在变换器1之后，进行空间池化，如图1中rvt所示。变换器1与2中分别包含12个与4个RVT-Block。为了匹配解码器输出维度，对变换器2的输出进行卷积及上采样。为了解决随着网络的加深，模型准确度饱和以及性能退化的问题，将rvt子模块的输入与上采样后的特征图相加，得到rvt的输出。

图 2 变换器模块示意图

下载: 全尺寸图片幻灯片

2.2.1 特征图预处理

本文采用卷积嵌入的方法进行特征图重构。特征图预处理过程如图3所示。首先对输入的特征图进行卷积操作再归一化。然后通过平均池化提取特征图的局部信息，同时通过最大池化提取特征图峰值信息。

图 3 特征图预处理示意图

下载: 全尺寸图片幻灯片

将经过两种池化操作的特征图融合，再进行卷积操作。预处理的输出Y_pe可表示为

${{\boldsymbol{Y}}_{{\text{pe}}}} = {\text{Con}}{{\text{v}}_2}\left( {{\text{Pool}}\left( {{\text{BN}}\left( {{\text{Con}}{{\text{v}}_{\text{5}}}\left( {{{\boldsymbol{X}}_{{\text{pe}}}}} \right)} \right)} \right)} \right)$

(1)

其中， ${{\boldsymbol{X}}_{{\text{pe}}}}$ 为输入特征图， ${\text{Con}}{{\text{v}}_i}\left( {{\cdot}} \right)$ 表示i×i卷积操作， ${\text{BN}}\left( {{\cdot}} \right)$ 表示归一化， ${\text{Pool}}\left( {{\cdot}} \right)$ 表示池化。

2.2.2 变换器(Transformer)

本文提出变换器子模块根据像素间的关系提取预处理后特征图的全局特征，其结构如图2所示。受自注意力机制^[14]的启发，本文将多个通道同时进行自注意力计算，如图4所示。把尺度 $\left( {c,h,w} \right)$ 的特征图压缩维度后变成 $\left( {h \times w,c} \right)$ ，再变成 $\left( {6,h \times w,c/6} \right)$ ，本文对x₁～x₆同时进行自注意力机制计算，这大大提高了特征的提取速度。以全连接的方式将3个可训练权重W_Q, W_K和W_V分别作用于每个通道中的特征图x_i，得到对应的q (query),k (key)和v (value) 3个值。利用q和k的乘积表示特征图的相关性，为了利于网络训练，需要对q和k的乘积做线性缩放，缩放因子为 $\sqrt {{c \mathord{\left/ {\vphantom {c 6}} \right. } 6}}$ 。w_i,j经Softmax操作得到对应位置的注意力权重A_i,j ，再与相应的v_i相乘，得到一个通道下的注意力结果b_i，最后将6个通道的注意力结果以通道维度拼接。这里的注意力机制可表示为式(2)。变换器对特征图处理时，将输入的特征图先压缩维度，经过重新排列(Rearrange)把形状为 $\left( {c,h,w} \right)$ 的特征图变为 $\left( {h \times w,c} \right)$ 。压缩后的2维序列 ${{\boldsymbol{X}}_{{\text{RB}}}}$ 作为RVT-Block模块的输入，RVT-Block模块的输出记为 ${{\boldsymbol{Y}}_{{\text{RB}}}}$ 。 ${{\boldsymbol{Y}}_{{\text{RB}}}}$ 与 ${{\boldsymbol{X}}_{{\text{RB}}}}$ 的关系如式(3)所示

图 4 多通道自注意力示意图

下载: 全尺寸图片幻灯片

${\text{attention}} = {\text{Softmax}}\left( {\frac{{{{\boldsymbol{Q}}_{\text{I}}}{{\boldsymbol{K}}_{\text{I}}}^{\text{T}}}}{{\sqrt {{{{c}} \mathord{\left/ {\vphantom {{{c}} {\text{6}}}} \right. } {\text{6}}}} }}} \right){{\boldsymbol{V}}_{\text{I}}}$

(2)

${{\boldsymbol{Y}}_{{\text{RB}}}} = D\left( {{\text{mlp}}\left( {{\text{LN}}\left( {{{\boldsymbol{X}}_1}} \right)} \right)} \right) \oplus D\left( {{\text{att}}{{\text{n}}_{{\text{RB}}}}\left( {{{\boldsymbol{X}}_{{\text{RB}}}}} \right)} \right) \oplus {{\boldsymbol{X}}_{{\text{RB}}}}$

(3)

其中， ${\text{att}}{{\text{n}}_{{\text{RB}}}}\left( {{\cdot}} \right)$ 表示图4所示的注意力机制， $D\left( {{\cdot}} \right)$ 表示失活， $\oplus$ 表示像素级相加， ${\text{LN}}\left( {{\cdot}} \right)$ 表示归一化， ${\text{mlp}}\left( {{\cdot}} \right)$ 表示多层感知。经过L个RVT-Block模块运算得到2维特征图，最后将特征图恢复到(c,h,w)。为了解决随着网络加深，模型性能退化的问题，本文使用了残差机制；为了更准确地检测图像显著性在模型中加入了多层感知机(MultiLayer Perceptron, MLP)，通道扩张率设为4。

2.3 MA模块

本文提出空间注意力模块(Spatial Attention Module, SAM)和通道注意力模块(Channel Attention Module, CAM)，首先从空间和通道两个维度提取特征，再对SAM设置加权因子 $\beta$ ，CAM设置加权因子 ${{1 - }}\beta$ 。两者求和的结果经1×1卷积有选择地融合显著信息，如图1中MA模块所示。 $\beta \in \left[0, 0.2,0.4,0.5,0.6,0.8,1.0 \right]$ ，本文将在实验部分给出SAM与CAM的最佳比重。

2.3.1 SAM

SAM子模块负责获得空间注意力，示意图如图5所示。输入尺度为(c ,h ,w)的特征图，经过3次 $1 \times 1$ 卷积，将3个不同的可学习权重作用于特征图，并压缩维度(Rearrange)，便于计算空间中任一特征像素与其他像素之间的关系，得到尺度为 $\left( {c,h \times w} \right)$ 的query值Q_s、key值K_s和value值V_s。在特征图的空间维度有 $h \times w$ 个像素，计算所有像素两两之间的注意力关系得到 $\left( {h \times w,h \times w} \right)$ 的空间注意力得分矩阵，即 $\dfrac{{{\boldsymbol{Q}}_{\text{S}}^{\text{T}}{{\boldsymbol{K}}_{\text{S}}}}}{{\sqrt c }}$ ， $\sqrt c$ 为缩放因子。特征图V_s的像素与对应位置的注意力权重相乘，即 ${{\boldsymbol{V}}_{\text{S}}} \times {\bf{Att}}$ ，得到空间注意力结果。然后将注意力结果经线性变换再提升维度，加入多层感知机MLP，通道扩张率设为4。所得结果与输入特征图加和，解决了模型性能退化的问题。计算流程用公式可表示为

图 5 SAM示意图

下载: 全尺寸图片幻灯片

${{\boldsymbol{Y}}_{\text{S}}} = {\text{mlp}}\left( {\varOmega \left( {{\text{att}}{{\text{n}}_{\text{S}}}\left( {{\text{Con}}{{\text{v}}_{\text{S}}}\left( {{{\boldsymbol{X}}_{\text{S}}}} \right)} \right)} \right)} \right) \oplus {{\boldsymbol{X}}_{\rm{S}}}$

(4)

其中， ${{\boldsymbol{X}}_{\text{S}}}$ 表示SAM模块的输入图像， ${\text{Con}}{{\text{v}}_{\text{S}}}\left( {{\cdot}} \right)$ 表示1×1卷积与重新排列操作， ${\text{att}}{{\text{n}}_{\text{S}}}\left( {{\cdot}} \right)$ 表示SAM模块的空间注意力机制， $\varOmega \left( {{\cdot}} \right)$ 操作由线性变化、失活以及重新排列组成， ${\text{mlp}}\left( {{\cdot}} \right)$ 表示多层感知， $\oplus$ 表示像素级相加。

2.3.2 CAM

CAM子模块用来获得通道注意力，如图6所示。将尺寸为(c ,h ,w)的输入特征图经过球形卷积，再压缩(Flatten)成 $\left( {c,h \times w} \right)$ 的尺寸。对特征矩阵进行转置(Transpose)，再将一组尺寸为 $\left( {h \times w,c} \right)$ 的可学习参数加到转置后的特征图上进行位置编码^[15]。以相同的固定权重得到query值Q_c、key值K_c和value值V_c，计算特征图通道之间的注意力关系， $\sqrt {hw}$ 为缩放因子。将所得通道注意力得分矩阵与V_c相乘，加入通道扩张率为4的MLP。串联4个CAM-Block(L_c=4)。使用残差机制，得到CAM的输出。CAM-Block的计算流程如式(5)所示

图 6 CAM示意图

下载: 全尺寸图片幻灯片

${{\boldsymbol{Y}}_{\text{C}}} = \varPhi \left( {{\text{att}}{{\text{n}}_{\text{C}}}\left( {{\text{LN}}\left( {{{\boldsymbol{X}}_{\text{C}}}} \right)} \right)} \right) \oplus {\text{mlp}}\left( {\varPhi \left( {{\text{att}}{{\text{n}}_{\text{C}}}\left( {{\text{LN}}\left( {{{\boldsymbol{X}}_{\text{C}}}} \right)} \right)} \right)} \right)$

(5)

其中， ${\text{LN}}\left( \cdot \right)$ 表示归一化， ${\text{att}}{{\text{n}}_{\text{C}}}\left( \cdot \right)$ 表示CAM模块的通道注意力机制， $\varPhi \left( \cdot \right)$ 表示由归一化、失活以及转置组成的操作， ${\text{mlp}}\left( \cdot \right)$ 表示多层感知， $\oplus$ 表示像素级相加。

2.4 纬度加权损失函数

本文设计一种新的损失函数，使赤道部分的损失权重最高，向两极的损失权重递减。由于同一纬度的像素失真程度相同，随着纬度的变化，相应纬度位置的像素其失真程度也发生变化，因此以像素点的纬度值来确定不同位置像素的损失权重。设全景图像的高为h，宽为w，把整幅图像分成h个纬度带。则第i个纬度带的损失权重 ${W_i}$ 与损失函数Loss可表示为

${W_i} = {\text{sin}}\left( {\frac{{i{\pi}}}{h}} \right) + 1 , {\text{Loss}} = \frac{1}{N}\sum\limits_{i,j} {{W_i}{{\left( {{S_{i,j}} - {G_{i,j}}} \right)}^2}}$

(6)

其中， ${S_{i,j}}$ 表示预测的显著图中位于第i个纬度带的第j个像素灰度值， ${G_{i,j}}$ 表示基准显著图中位于第i个纬度带的第j个像素的灰度值，N表示图像的像素个数。

3. 实验

3.1 数据集与预处理

AOI有600张全景图像和基准显著图。Salient360有85张全景图像和基准显著图。由于原始数据集的图像数量较少，本文采用垂直翻转、水平翻转和双向翻转进行数据增强。此外，基于噪声的数据增强，能够提高模型在全景图像包含噪声情境下的鲁棒性能^[5]，因此本文对全景图像添加了高斯、泊松、椒盐和斑点噪声。上述所有的数据增强操作，不会改变全景图像像素之间的依赖关系，因此不会影响图像的显著性。最终构建的AAOI, ASalient360数据集分别包含4800张、680张全景图像。

3.2 实验设置

本文使用Pytorch架构以及150个epoch训练模型，4张图像为一组进行批处理。学习率设置为0.0001，动量参数为0.9的随机梯度下降(Stochastic Gradient Descent, SGD)优化器，权重衰减0.00001。设置检查点，参考文献[5]中训练图像与测试图像的比例，AAOI数据集中4752张图像用于训练，48张图像用于测试；ASalient360数据集中640张图像用于训练，40张图像用于测试。训练集中训练部分与验证部分图像数量比例为17:3。本文所有实验均在配备RTX3060 GPU和AMD I5 3600 CPU的台式机上完成。

3.3 MA模块中 $\beta$ 取值对模型性能的影响

表1列出了MA模块中SAM与CAM子模块不同权重的实验结果。Grade如式(7)所示

表 1 不同加权因子的实验结果

SMA $\beta$	CAM ${{1 - } }\beta$	CC↑	SIM↑	KLDiv↓	NSS↑	AUC_Judd↑	AUC_Borji↑	Grade
0	1.0	0.9005	0.7787	0.1970	3.5346	0.9914	0.9755	3.3809
0.2	0.8	0.8803	0.7785	0.4403	3.4964	0.9861	0.9618	0.7047
0.4	0.6	0.9008	0.7921	0.5070	3.7190	0.9936	0.9795	3.4726
0.5	0.5	0.9067	0.8119	0.2198	3.2849	0.9898	0.9772	3.5050
0.6	0.4	0.8912	0.7871	0.2350	3.2212	0.9840	0.9655	1.2312
0.8	0.2	0.8771	0.7561	0.2317	3.2893	0.9864	0.9756	1.1757
1.0	0	0.9023	0.7890	0.3775	3.6737	0.9919	0.9774	3.4871

下载: 导出CSV

| 显示表格

$\begin{split} {\text{Grade}} =& {\text{CC}} + {\text{SIM}} - {\text{KLDiv}} + {\text{NSS}} \\ & + {\text{AUC\_Judd}} + {\text{AUC}}\_{\text{Borji}} \end{split}$

(7)

其中，CC,SIM,KLDiv,NSS,AUC_Judd以及AUC_Borji分别是6种评价指标标准化后的结果。由式(7)可知，Grade同等考虑了6个评价指标的作用，得分越高，模型的性能越好。由表1中Grade结果可知，当 $\beta$ =0.5时，即SAM与CAM具有相同权重时，模型性能在列出的所有权重组合中达到最优。本文模型取 $\beta$ =0.5。

3.4 实验对比

本文将URMNet模型与目前先进的方法进行比较，包括U型网络(U-Net)^[16]、注意力U型网络(Attention U-Net, AttU-Net)^[17]、球形U型网络(Spherical U-Net)^[3]、全景卷积神经网络(panoramic Convolutional Neural Network, panoramic CNN)^[5]、四方注意力U型网络(Quartet Attention U-Net, QAU-Net)^[18]以及从通道和变换器角度考虑U-Net的跳跃连接网络(rethinking the skip connections in U-Net from a Channel-wise perspective with transformer, UCTransNet)^[19]。不同对比方法都是在相同的实验环境中运行测试。

表2给出了6个评价指标在AAOI数据集上的定量比较结果。明显可知，6个评价指标下，本文模型的性能超过所有对比方法的性能。具体来说，本文模型在指标相关系数(Correlation Coefficien, CC)、相似度(SIMilarity, SIM)、KL散度(KL-Divergence, KLDiv)、标准化扫描路径显著性(Normalized Scanpath Saliency, NSS)以及分别经Judd和Borji优化后的ROC曲线下面积(Area Under the Curve,AUC)AUC_Judd和AUC_Borji上，与对比的6种优秀算法中最好的指标(表2中用红色标出)相比，优化幅度分别是3.13%, 2.42%, 17.80%, 11.75%, 0.52%和0.68%。表3列出了6个评价指标在ASalient360数据集上的定量比较结果。本文模型在指标CC, SIM, NSS, AUC_Judd和AUC_Borji上，与对比的6种优秀算法中最好的指标(表3中用红色标出)相比，优化幅度分别是7.36%, 1.15%, 8.41%, 0.69%和1.67%。在KLDiv指标上，本文模型比文献算法最佳结果低0.1785。这是由于为了使综合性能更优，本文模型在设计上平衡了准确性、泛化性以及复杂度。总地来说，两个数据集下的实验结果证明了本文所提出模型的先进性。

表 2 AAOI数据集上各模型客观指标对比

方法	CC↑	SIM↑	KLDiv↓	NSS↑	AUC_Judd↑	AUC_Borji↑
URMNet	0.8934	0.7918	0.1787	3.7113	0.9865	0.9707
U-Net(2015)^[16]	0.8550	0.7694	0.2647	2.9639	0.9741	0.9582
AttU-Net(2018)^[17]	0.8663	0.7675	0.2359	3.3212	0.9796	0.9632
Spherical U-Net(2018)^[3]	0.7832	0.7304	0.3167	2.4795	0.9467	0.9295
panoramic CNN(2020)^[5]	0.8520	0.7533	0.2412	3.1999	0.9778	0.9641
QAU-Net(2021)^[18]	0.7314	0.6530	0.4203	1.8678	0.9226	0.8926
UCTransNet(2021)^[19]	0.8619	0.7731	0.2105	3.0204	0.9814	0.9625

下载: 导出CSV

| 显示表格

表 3 ASalient360数据集上各模型客观指标对比

方法	CC↑	SIM↑	KLDiv↓	NSS↑	AUC_Judd↑	AUC_Borji↑
URMNet	0.6683	0.6602	0.5834	2.9874	0.9449	0.9336
U-Net(2015)^[16]	0.6061	0.6404	0.4397	2.6228	0.9180	0.8973
AttU-Net(2018)^[17]	0.5589	0.6262	0.4892	1.9840	0.8871	0.8644
Spherical U-Net(2018)^[3]	0.6028	0.6412	0.6714	2.5834	0.9384	0.9183
panoramic CNN(2020)^[5]	0.6225	0.6527	0.4049	2.0324	0.8975	0.8482
QAU-Net(2021)^[18]	0.5641	0.6280	0.5138	2.7556	0.8889	0.8859
UCTransNet(2021)^[19]	0.5429	0.6165	0.5246	2.0041	0.8990	0.8683

下载: 导出CSV

| 显示表格

为了进一步验证本文模型的性能，与6种先进方法进行可视化对比，结果如图7和图8所示。图9是为了便于观察，将图7第2行放大。结果表明，本文方法能够较为准确地检测出图像的显著性区域，包括远景全景图像(图7第1,3,6行)和近景全景图像(图7第2,4,5行，图8第4,5行)。远景图像中图7第6行，背景为图像左侧的大树以及蓝色天空，显著区域为图像中间红色树以及右侧白色天空对应区域，从结果上看，本文算法更好地抑制了背景对显著性区域检测的干扰，使得检测结果与人工标注的真实值之间更为接近。近景图像中图7第5行，显著性区域有3处，分别是图像中间的雕像以及雕像两边的书法作品所在区域。对比模型由于背景的干扰，检测的显著性区域明显大于真实值。本文模型的检测结果有效地改善了这一现象，使显著性区域定位更准确。另外，本文模型对噪声干扰(图7第2行，图8第2行)更为鲁棒。图9中显著区域包括图像中间较大的两幅画所在区域，以及包括所有画在内左右方向的带状区域。受噪声干扰，对比模型中U-Net, AttU-Net, Spherical U-Net以及QAU-Net，检测的显著性区域明显大于真实值。与检测较为准确的panoramic CNN和UCTransNet相比，本文模型检测结果更接近于真实值。

图 7 本文方法与其他方法在AAOI数据集上的可视化对比

下载: 全尺寸图片幻灯片

图 8 本文方法与其他方法在ASalient360数据集上的可视化对比

下载: 全尺寸图片幻灯片

图 9 添加噪声后的显著性检测图像放大可视化对比

下载: 全尺寸图片幻灯片

3.5 消融实验

3.5.1 URMNet网络和损失函数Loss的消融实验

表4中，Baseline表示具有球形卷积的U-Net网络、L₁表示文献[10]提出的自适应加权损失函数、L₂表示文献[20]提出的固定权重损失函数、L₃表示结合文献[6]和文献[21]给 ${\text{CC}}$ , ${\text{NSS}}$ 和 ${\text{KLDiv}}$ 相同比例权重构造的损失函数。将两种网络(Baseline和URMNet)与4种损失函数(L₁,L₂,L₃和Loss)分别组合进行消融实验，实验结果如表4所示。可知URMNet与Loss组合时，模型性能最好，说明采用损失函数Loss训练URMNet网络对获得最佳显著性检测性能是必要的和有效的。

表 4 网络和损失函数的消融实验

模型		损失函数				CC↑	SIM↑	KLDiv↓	NSS↑	AUC_Judd↑	AUC_Borji↑
Baseline	URMNet	L₁	L₂	L₃	Loss	CC↑	SIM↑	KLDiv↓	NSS↑	AUC_Judd↑	AUC_Borji↑
√		√				0.7389	0.6935	2.2079	1.5107	0.8634	0.8310
√			√			0.7300	0.6886	2.2614	1.5002	0.8636	0.8169
√				√		0.6629	0.6604	1.2939	1.2369	0.8184	0.7806
√					√	0.8604	0.7618	0.3906	3.0463	0.9846	0.9614
	√	√				0.7456	0.7007	2.2068	1.5043	0.8627	0.8304
	√		√			0.7565	0.6845	2.9497	1.6848	0.8909	0.8166
	√			√		0.7557	0.6909	3.1000	1.6607	0.8830	0.8465
	√				√	0.9067	0.8119	0.2198	3.2849	0.9898	0.9772

下载: 导出CSV

| 显示表格

3.5.2 RVT模块和MA模块的消融实验

实验结果如表5所示。同时加入RVT和MA模块，6个性能指标较表4第1行指标增幅分别为7.0%,5.6%,22.7%,11.7%,1.3%和1.96%，能够明显提高各项性能指标得分，增强网络显著性检测性能，这表明同时加入RVT和MA模块对本文模型获得最佳显著性检测效果是必要的和有效的。

表 5 RVT和MA模块消融实验结果

RVT	MA	CC↑	SIM↑	KLDiv↓	NSS↑	AUC_Judd↑	AUC_Borji↑
×	×	0.8335	0.7392	0.2947	2.9639	0.9779	0.9604
√	×	0.8664	0.7788	0.3682	3.0885	0.9821	0.9702
×	√	0.8553	0.7118	0.3317	3.4291	0.9848	0.9576
√	√	0.8922	0.7805	0.2278	3.3100	0.9910	0.9786

下载: 导出CSV

| 显示表格

3.6 泛化性能对比实验

为了检测本文模型的泛化能力，与优秀模型进行泛化性能对比，结果如表6所示。模型在AAOI数据集上进行训练，在ASalient360数据集上进行测试，可以看出URMNet在指标CC, SIM, NSS, AUC_Judd以及AUC_Borji上，较优秀算法中最好指标(红色标出)优化幅度分别为9.04%, 2.02%, 5.97%, 0.89%和0.48%。说明URMNet模型在保证显著性检测精度前提下，有较好的泛化能力。

表 6 不同模型泛化性能对比

方法	CC↑	SIM↑	KLDiv↓	NSS↑	AUC_Judd↑	AUC_Borji↑
URMNet	0.5899	0.6116	1.0708	1.8395	0.9181	0.8917
U-Net(2015)^[16]	0.5402	0.5733	1.9622	1.6975	0.9022	0.8750
AttU-Net(2018)^[17]	0.4906	0.5572	1.9561	1.4668	0.8896	0.8556
panoramic CNN(2020)^[5]	0.5146	0.5770	1.1245	1.4486	0.8865	0.8504
QAU-Net(2021)^[18]	0.5044	0.5989	0.5373	1.5522	0.8854	0.8362
UCTransNet(2021)^[19]	0.5410	0.5760	1.8189	1.7359	0.9100	0.8874

下载: 导出CSV

| 显示表格

3.7 复杂度对比实验

复杂度实验结果如表7所示。可知URMNet计算量较U-Net, QAU-Net和UCTransNet优化幅度分别为62.35%, 96.07%和97.24%。参数量较QAU-Net和UCTransNet优化幅度分别为34.28%和60.93%。表明本文模型在保证较高检测精度的同时，相对较好地控制了模型的复杂度。

表 7 不同模型复杂度对比

复杂度指标	U-Net^[16]	QAU-Net^[18]	UCTransNet^[19]	URMNet
GFLOPs(G)	4.1345	39.5893	56.4113	1.5567
Params(M)	23.2655	41.8559	70.4118	27.5091

下载: 导出CSV

| 显示表格

4. 结束语

针对当前全景图像显著性检测方法存在检测精度偏低、模型收敛速度慢和计算量大的问题，本文提出一种基于鲁棒视觉变换和多注意力的U型网络URMNet，用于全景图像的显著性检测。通过RVT用来提取4种尺度特征图所包含的显著信息，MA模块融合多维度注意力，提高中间层特征提取能力，纬度加权损失函数Loss提高检测精度并减少模型训练时间，使用球形卷积，最大限度上降低图像失真。实验结果表明，该模型收敛速度快，检测结果精度高于目前主流方法的检测结果精度，同时该模型具有较好的泛化性能和较低的复杂度。后续工作将进一步优化全景图像显著性检测方法，提高显著性检测效率。

图 1 空中交通管理系统资产

下载: 全尺寸图片幻灯片

图 2 空中交通管理系统信息交换模型

下载: 全尺寸图片幻灯片

图 3 综述框架

下载: 全尺寸图片幻灯片

图 4 民机航电网络内部资产^[31]

下载: 全尺寸图片幻灯片

图 5 民机航电网络内部架构及外部连接关系^[11]

下载: 全尺寸图片幻灯片

图 6 民用飞机内部连接^[27]

下载: 全尺寸图片幻灯片

图 7 通过ACARS攻击FMS的ATT&CK模型

下载: 全尺寸图片幻灯片

图 8 CNS/ATM空管系统信息流图^[70]

下载: 全尺寸图片幻灯片

图 9 基于ATT&CK的ATM攻击分析

下载: 全尺寸图片幻灯片

图 10 空管SWIM总体架构^[12]

下载: 全尺寸图片幻灯片

图 11 机场信息流图

下载: 全尺寸图片幻灯片

图 12 基于ATT&CK的机场和航空公司的攻击分析

下载: 全尺寸图片幻灯片

图 13 不同飞行阶段的智能化空管典型应用^[149]

下载: 全尺寸图片幻灯片

图 14 基于ATT&CK智能化空管的攻击分析

下载: 全尺寸图片幻灯片

图 15 空管系统ATT&CK矩阵

下载: 全尺寸图片幻灯片

表 1 空中交通管理系统信息安全保障框架

国家/组织	制定时间	框架名称
美国	1999	Information Systems Security
日本	2010	Collaborative Actions for Renovation of Air Traffic Systems
中国	2012	民用航空网络与信息安全管理规范
欧盟	2013	Global ATM Security Management
德国	2015	Air Traffic Resilience
美国CSFI	2015	Air Traffic Control Cyber Security Project

下载: 导出CSV

表 2 现有综述文献总结

文献	资产梳理			信息安全
文献	飞机	ATM	机场	传统安全	智能安全	动态安全
[11]	√	√		√		√
[23]			√		√
[24]			√	√
[25]			√	√
[26]			√	√	√
[27]	√			√
[28]			√	√
[29]			√	√
[30]		√		√
[12]	√	√	√	√
本文	√	√	√	√	√	√

下载: 导出CSV

表 3 航电网络各管理域相关资产梳理

管理域	ACD^[39–44]	AISD^[45–47]	乘客域^[48,49]
资产	客舱服务系统(CSS) 飞行管理系统(FMS) 自动驾驶系统飞行数据记录系统(FDR) 驾驶舱语音记录器(CVR) 电子飞行仪表系统(EFIS) 发动机警报系统(EICAS)	客舱管理终端(CMT) 客舱管理计算机(CMC) 文件服务模块(FSM) 驾驶舱打印机(CPPT) 电子飞行包(EFB) 客舱广播及内话系统(PACIS)	机上娱乐和通信系统(IFEC) 乘客娱乐终端

下载: 导出CSV

表 4 民用飞机面临的具体攻击样式及分类

资产	攻击技术	攻击影响	STRIDE
飞行管理系统^[55–57]	向FMS注入虚假数据	操纵控制装置	篡改
	用损坏的导航数据更新FMS
	远程更新恶意的航空公司修改信息(AMI)
发动机警报系统^[58]	注入恶意的发动机警报消息	安全性损失	欺骗
机上娱乐和通信系统^[59]	获取乘客的显示系统的控制权	播种恐惧	篡改
	获取客舱的控制权	播种恐惧	权限提升
	收集乘客的隐私信息	盗窃乘客信息	信息泄露
电子飞行包^[60]	利用蓝牙、Wi-Fi或蜂窝网络获取EFB的控制权	失去控制	篡改
	通过检索日志收集应用程序的存储信息	盗窃飞机统计数据	信息泄露
	修改飞行计划和导航数据	操纵控制装置	篡改
	安装恶意应用程序	操纵控制装置	篡改
	滥用托管应用程序并修改操纵指令(例如控制客舱灯光)	播种恐惧	权限提升

下载: 导出CSV

表 5 空管通信、导航、监视系统资产^[68,70,71]

	资产名称	用途	发送方	接收方
通信系统	CPDLC	数字通信	飞机和地面站	飞机和地面站
	SATCOM	卫星通信	卫星、飞机和地面站	飞机和地面站
	ACARS	数字通信	飞机和地面站	飞机和地面站
导航系统	DME	测量距离	地面站	飞机
	VOR	提供方位	地面站	飞机
	ILS	着陆引导	地面站	飞机
	GNSS	定位和导航	卫星和地面站	飞机
监视系统	PSR	提供位置、速度和航行等信息	地面站	飞机和地面站
	SSR	提供识别码、高度、速度等信息	飞机	飞机和地面站
	ADS-B	提供位置、速度和航行等信息，防碰撞	卫星和飞机	飞机和地面站
	MLAT	提供位置信息	飞机	地面站

下载: 导出CSV

表 6 空管通信、导航、监视系统资产面临的安全威胁

资产	攻击技术	攻击影响	STRIDE
CPDLC^[87]	用无线电接收机接收、解析和收集信息；注入定制的虚假消息；篡改飞机和空中交通服务机构之间的通信；洪泛攻击	窃取飞机相关信息；欺骗飞行员和管制员；飞行员无法获取有效通信数据	信息泄露；欺骗；篡改；拒绝服务
SATCOM^[88]	破坏、拦截或修改机上Wi-Fi；破坏飞行和卫星的通信	乘客和机组人员无法获取服务	拒绝服务
ACARS^[89]	飞机侦察；发布并传输虚假的信息	飞机航迹泄露；扰乱空管态势	信息泄露；欺骗
DME/VOR/ILS^[90]	飞机侦察；使用定制信号压制合法信号	飞机航迹泄露；破坏飞行安全	信息泄露；欺骗
GNSS^[91]	使用GPS信号欺骗GNSS系统；发送比卫星 GPS信号强的信号	破坏飞行安全；扰乱空管态势	欺骗；篡改；拒绝服务
PSR/SSR^[92]	攻击者使用软件无线电修改、阻碍、注入监视雷达消息	破坏空管态势	欺骗；篡改；拒绝服务
ADS-B^[93]	飞机侦察；幽灵飞机注入；轨迹篡改；泛洪拒绝	飞机航迹泄露；扰乱空管态势；扰乱空管态势	信息泄露；欺骗；篡改；拒绝服务
MLAT^[94]	阻碍GPS信号的信号同步；使用多个设备欺骗GPS信号	扰乱空管态势	欺骗

下载: 导出CSV

表 7 针对空管通信、导航、监视系统的安全防御手段

	资产	攻击	防御
通信系统	CPDLC	信息泄露；欺骗；篡改；拒绝服务	加密、身份认证^[98]
	SATCOM	拒绝服务	物理层安全^[99]
	ACARS	信息泄露；欺骗	加密、身份认证^[100]
导航系统	DME/VOR/ ILS	信息披露、欺骗	加密、身份认证^[101]
导航系统	GNSS	欺骗；篡改；拒绝服务	入侵检测、加密^[102]
监视系统	PSR/SSR	欺骗；篡改；拒绝服务	入侵检测^[103]
	ADS-B	信息泄露；欺骗；篡改；拒绝服务	入侵检测、物理层指纹、加密^[80,84–87]
	MLAT	欺骗	消息验证^[104]

下载: 导出CSV

表 8 机场资产总结

	名称		用途
机场管理	企业管理系统		用于机场的行政管理
	资产盘点系统		跟踪和管理机场的物理资产
	人力资源系统		管理机场员工的招聘、培训、调度和绩效评估
空侧运营	空中交通管理系统		指挥和监控飞机在空中的飞行路径
	导航着陆设备		辅助飞机安全起降
	飞行追踪系统		实时监控飞机位置
	气象信息系统		提供实时气象数据
	出发控制系统		管理飞机的起飞顺序和时间
	航空运营中心		处理机场和航空公司的业务
	货物处理系统		管理机场的货物装卸、存储和运输
	通信导航监视系统		集成通信、导航和监视功能
	机场运行数据库		存储和管理机场运营的关键数据
陆侧运营	机场陆侧运营系统控制中心		协调机场地面交通和运营活动
	自动车辆识别系统		自动识别和监控机场内的车辆
	燃料管理系统		管理机场的燃料供应和分配
	交通运输系统		提供机场与城市之间的交通连接
	寻路服务系统		帮助乘客在机场内导航
安全和安保	身份验证系统		验证乘客和员工的身份
	行李处理系统		自动化处理乘客行李的检查、分拣和运输
	智能监控系统		监控机场的关键区域
	海关和移民		管理国际旅客的入境和出境检查
	简易爆炸物探测系统		检测潜在的爆炸物威胁
	门禁系统		控制对机场敏感区域的访问
	周界入侵检测系统		监测机场周边
	应急响应系统消防系统		在紧急情况下提供快速响应
IT和通信	内部	局域网和虚拟私人网络	为机场内部提供安全的网络连接
		通信系统(如无线电频谱管理系统)	为机场提供通信服务
		存储数据	保存机场运营中产生的数据
		IT设备(软件和硬件)	支持机场的信息技术需求
		飞行显示管理系统	向乘客提供航班信息
		移动电话网络或APP	提供移动通信服务
	外部	全球定位系统	提供精确的定位服务
		网络安全管理	保护机场网络不受攻击
		广域网	连接机场与外部网络
		地理信息系统	用于机场的地理数据分析和地图服务
设施与维护	机场车辆维修		维护和修理机场车辆
	能源管理(发电机)		管理机场的能源供应
	数据采集与监视控制系统		监控和控制机场的关键基础设施
乘客系统	后勤系统		支持机场的后勤服务
	电子显示屏		机场内显示信息
	乘客办理登机和登机系统		自动化处理乘客的登机手续
	乘客姓名登记系统		管理乘客信息
	预约系统		管理航空公司的座位预订和航班安排
	行李处理系统		自动化处理乘客行李的检查、分拣和运输

下载: 导出CSV

表 9 机场资产面临的安全威胁

资产	技术	攻击面	影响	STRIDE
机场管理	DDoS攻击	有线通信与IT资产的交互	机场服务器瘫痪	拒绝服务
	滥用授权	物理与IT资产的交互	未授权人员访问敏感系统	权限提升
	社交和网络钓鱼攻击	与员工或乘客的交互	诱导员工泄露敏感信息	信息泄露
空侧运营	通信攻击	无线通信与IT资产的交互	通信系统被监听或篡改	篡改
	恶意软件	资产间的跳跃攻击	数据丢失或被远程控制	欺骗
	滥用授权	物理与IT资产的交互	未授权人员访问敏感系统	权限提升
陆侧运营	网络攻击	有线通信与IT资产的交互	监控设施或注入受损数据	欺骗
	篡改设备	有线通信与IT资产的交互	设备的功能和数据被修改	篡改
	社交和网络钓鱼攻击	与员工或乘客的交互	诱导员工泄露敏感信息	信息泄露
安全和安保	恶意软件	资产间的跳跃攻击	数据丢失或被远程控制	欺骗
安全和安保	滥用授权	物理与IT资产的交互	未授权人员访问敏感系统	权限提升
IT和通信	DDoS攻击	有线通信与IT资产的交互	机场服务器瘫痪	拒绝服务
	通信攻击	无线通信与IT资产的交互	通信系统被监听或篡改	篡改
	恶意软件	资产间的跳跃攻击	数据丢失或被远程控制	篡改
	网络攻击	有线通信与IT资产的交互	监控设施或注入受损数据	欺骗
	社交和网络钓鱼攻击	与员工或乘客的交互	诱导员工泄露敏感信息	信息泄露
设施与维护	滥用授权	物理与IT资产的交互	未授权人员访问敏感系统	权限提升
	网络攻击	有线通信与IT资产的交互	监控设施或注入受损数据	欺骗
	社交和网络钓鱼攻击	与员工或乘客的交互	诱导员工泄露敏感信息	信息泄露
乘客系统	篡改设备	有线通信与IT资产的交互	设备的功能和数据被修改	篡改
乘客系统	恶意软件	资产间的跳跃攻击	数据丢失或被远程控制	欺骗

下载: 导出CSV

表 10 机场资产的安全防御手段

攻击	涉及资产	安全防御
DDoS攻击	企业管理系统	入侵检测/保护系统安全强化^[131,132] 防火墙、网络分段和纵深防御^[133,134] IT资产的灾难恢复计划^[135]
	人力资源系统
	IT设备(软件和硬件)
	网络安全管理
	广域网
	移动电话网络或APP
	网络安全管理
通信攻击	空中交通管理系统	入侵检测/保护反欺骗控制^[136,137] 强用户身份验证^[138] 应用安全和安全设计^[139,140] 数据加密^[141,142]
	导航着陆设备
	飞行追踪系统
	气象信息系统
	出发控制系统
	通信导航监视系统
	通信系统(如无线电频谱管理系统)
	飞行显示管理系统
	全球定位系统
	地理信息系统
恶意软件	通信系统(如无线电频谱管理系统)	入侵检测/保护反恶意软件技术控制 BYOD 最小特权访问管理^[143] 软件和硬件更新^[144] 防火墙、网络分段和纵深防御
	IT设备(软件和硬件)
	飞行显示管理系统
	全球定位系统
	广域网
	地理信息系统
	乘客姓名登记系统
	预约系统
篡改设备	自动车辆识别系统	限制外部设备的使用^[145] 入侵检测/保护数据加密增强物理安全和监控系统^[146]
	燃料管理系统
	交通运输系统
	寻路服务系统
	后勤系统
	电子显示屏
	乘客办理登机和登机系统
	乘客姓名登记系统
	预约系统
	行李处理系统
网络攻击	机场陆侧运营系统控制中心	防火墙、网络分段和纵深防御入侵检测/保护强用户身份验证更改设备的默认管理员凭据技术控制BYOD 数据加密
	自动车辆识别系统
	交通运输系统
	寻路服务系统
	通信系统(如无线电频谱管理系统)
	IT设备(软件和硬件)
	网络安全管理
	广域网
社交和网钓鱼攻击	企业管理系统	入侵检测/保护软件和硬件更新防火墙、网络分段和纵深防御反欺骗控制强用户身份验证应用程序安全和安全设计
	资产盘点系统
	人力资源系统
	IT设备(软件和硬件)
	移动电话网络或APP
	网络安全管理
	广域网
	数据采集与监视控制系统
滥用授权	企业管理系统	更改设备的默认凭据技术控制BYOD 软件和硬件更新最小特权和数据分类数据加密强用户身份验证用户访问管理
	资产盘点系统
	人力资源系统
	空中交通管理系统
	航空公司网关服务器系统
	货物处理系统
	机场运行数据库
	身份验证系统
	行李处理系统
	智能监控系统
	门禁系统
	周界入侵检测系统
	应急响应系统消防系统
	机场车辆维修
	能源管理(发电机)
	数据采集与监视控制系统

下载: 导出CSV

表 11 空管典型应用

飞行阶段	典型应用	模型类别	模型优势
起飞前阶段	FDP^[150–154]	图神经网络	基于图神经网络具有能够捕捉复杂关系、动态信息建模、上下文信息利用、提高泛化能力、多模态融合、高效计算的优势。
起飞前阶段	ARO^[155]	强化学习	基于强化学习的动态决策、长远优化、自适应学习、资源优化、高维度和复杂的调度、在线学习能力、低人工依赖优势。
降落阶段	ARR^[156–159]	目标检测算法	基于深度学习的跑道目标检测有检测实时性、适应复杂天气环境、人工干预降低、多源信息融合、导航能力、运营效率优化、多机协同、数据收集与分析优势。
降落阶段	ADS^[160–162]	目标检测算法	基于目标检测算法的反无人机技术在机场安全中具有实时监控与响应、精准检测、多传感器融合、自动化部署、提升事件响应效率、安全数据分析、协同作战、司法取证的优势。
起飞前阶段	APF^[163,164]	时序预测模型	时序预测模型在乘客数量预测中的应用，带来了历史数据利用、季节性波动感知、样本获取与自动更新、多变量出局处理、预测周期灵活性、决策效率、财务风险管理、成本效益、政策制定支持、客户体验优化优势。
飞行阶段	ATM^[165–169]	强化学习	基于强化学习的空中交通管理具有自适应决策、优化调度、实时反馈、多目标优化、解决高维问题、经验积累与迁移的优势。
飞行阶段	AFCF^[170–172]	时序预测模型	在航空燃油消耗预测中，时序预测模型具有动态数据处理、非线性关系获取、偶发事件处理、油量节约、历史趋势分析、多模型集成能力优势
起飞前阶段	FTP^[173–177]	时序预测模型	在航迹预测中，时序预测模型具有处理序列数据的能力、动态更新、非线性关系建模、自适应性、多变量输入、并行处理与效率、强大的特征提取优势。
飞行阶段	AWF^[178–182]	时序预测模型	在航空气象预测中，时序预测模型具有动态特性捕捉、气象延续性与记忆能力、自动特征学习、实时天气预测、气象预警功能、交互式决策支持优势。
降落阶段	AGVS^[183–186]	目标检测算法	在机场安全检测中，目标检测算法具有快速响应、高准确率、小目标检测能力、自动化监控、模型集成性强、复杂环境适应能力、可扩展性强、机场数据采集与分析、智能安全报警机制的优势。

下载: 导出CSV

表 12 攻击样式

攻击样式	模型类别	攻击优势
对抗样本攻击	图神经网络^[187–189]	(1)结构敏感性：GNN依赖于节点及其邻域之间的结构信息。对抗攻击可以通过小规模的边或节点操作来影响整个图的特征传播。(2)局部性与全局性的相互作用：GNN聚合邻居节点的信息，局部的扰动可能会引起全局的效果。(3)传播效应：图中某个节点的标签变化可以通过连接的边影响到其他节点，导致连锁反应，从而提高攻击效果。
	目标检测算法^[190,191]	(1)复杂性与高维特征：针对多维度特征，细微对抗扰动可以维持视觉效果的同时，降低模型的检测精度。(2)多阶梯输出：目标检测算法的决策过程涉及多步骤，每一阶段都可能受到对抗样本的影响。对抗攻击可针对特定阶段进行设计，从而提高攻击的成功率。(3)隐蔽性与实时性：对抗样本能够在视觉上保持自然状态，且易于实现针对现实世界中摄像头捕获的图像，这使得攻击具有较好的隐蔽性。
	时序预测模型^[192,193]	(1)序列长度与复杂性：时序数据长且具有复杂的时序关系。扰动可能对后续多个时刻的预测造成影响，使得攻击在序列中传播。(2)环境噪声与变化的脆弱性：算法经常需要应对环境噪声和随机性，对抗攻击可以精确地伪造这种噪声。
	强化学习^[194,195]	(1)环境动态性：强化学习通常在动态环境中进行决策。当攻击者在环境中施加对抗扰动时，这些扰动可以有效地影响智能体的观测状态和决策行为。(2)奖励信号：强化学习对抗样本可以通过改变环境状态，使得智能体接收到误导性的奖励信号。(3)多步决策链：强化学习攻击能在早期步骤干扰智能体的决策，将导致后续动作产生连锁反应，进一步降低整体任务表现。
模型窃取攻击	图神经网络^[196–198]	(1)信息聚合机制：这一机制意味着即便攻击者没有直接访问完整模型参数，也能通过观察输入-输出对推测出某些权重或结构特征，从而近似重建模型。(2)高维特征表达能力：GNN能够捕捉高维特征并能够灵活地处理不同大小和形状的图。(3)可互操作性与迁移：许多GNN模型是基于相似的原理和结构，攻击者可以从一个模型窃取知识，然后将其迁移到其他任务中。
后门攻击	图神经网络^[199,200]	(1)数据依赖性：GNN处理的输入数据通常包含节点及其连接关系，后门攻击可以通过注入特定的“干净”样本来劫持模型。(2)图拓扑结构利用：由于图拥有复杂的拓扑结构，攻击者可以设计特定的后门模式，使得在特定条件下影响某些节点或边的状态。
	目标检测算法^[201,202]	(1)后门隐匿性：攻击者可以通过在训练集中注入带有隐藏标记的图像样本，而这些图像貌似正常，因此更容易掩盖后门的存在。(2)标签操控：攻击者可以通过操控图像及其对应的标签，设计特定的后门触发条件。(3)难以回溯性：由于目标检测涉及大量的特征计算和复杂的数据流，后门的源头可能难以追踪，增加了后门被识别的难度。
	强化学习模型^[203,204]	(1)策略更新机制：攻击者可以设计一些特定的触发事件，引导代理在特定状态下意外地优化为不希望的策略，从而实现对模型的控制。(2)长期影响：由于RL侧重于长期奖励，攻击者可以精心选择后门样本，使得在触发后，代理会持续受到影响，从而实施潜在的危险行为。

下载: 导出CSV

表 13 智能化空管典型应用攻击样式的安全防御手段

攻击样式	安全防御	防御概述
对抗样本攻击	对抗训练	对抗训练在模型训练过程中，将对抗样本加入到训练集中。对于给定的训练样本，通过特定的攻击算法生成对抗样本，然后将也作为训练数据来更新模型参数。使模型在训练时就会接触到对抗样本，从而学习到对这些扰动更具鲁棒性的特征表示。
	输入预处理	(1)滤波技术去除对抗样本中的噪声扰动。它可以在一定程度上削弱添加的微小扰动，使对抗样本更接近原始样本，减少扰动对模型决策的影响。(2)特征压缩减少输入数据的维度或复杂度，去除可能包含对抗扰动的冗余信息，去除与对抗扰动相关的次要特征，提升模型鲁棒性。
	模型增强	(1)多模型融合是指将多个不同的模型进行融合，不同模型对对抗样本的脆弱性可能不同，通过融合可以综合各个模型的优势，提高整体的鲁棒性。(2)防御蒸馏借鉴知识蒸馏的思想，在防御对抗样本时，教师模型在对抗样本上的输出作为软标签，学生模型通过学习这些软标签来提高对对抗样本的鲁棒性。
	对抗样本检测	分析对抗样本和正常样本在统计特征上的差异，包括样本的激活值分布、梯度分布等统计量。对抗样本的这些统计量往往与正常样本不同，设定阈值可以检测出对抗样本。
模型窃取攻击	混淆技术	旨在通过对模型或模型输出进行各种变换，增加攻击者窃取模型知识的难度。
	访问控制与加密	对模型的参数、结构以及可能传输的数据进行加密处理。
	基于证明的防御	要求用户在获取模型的预测结果之前，完成一定的计算任务或提供某种形式的证明。
后门攻击	数据检测与清洗	(1)异常样本检测通过统计分析、聚类等方法识别训练数据中的异常样本。 (2)触发模式检测分析数据集中是否存在特定的触发模式。
	模型训练过程防御	在模型训练中使用 L1 和 L2 正则化等技术。减少后门攻击可利用的冗余参数，约束参数大小防止模型过拟合后门样本。
	模型检测与加固	(1)对训练好的模型进行逆向分析，检查模型结构和参数中是否存在隐藏的后门。(2)在模型训练过程中嵌入水印信息，通过检测水印的完整性来判断模型是否被篡改。
	运行时监测	(1)在模型运行时，实时监测输入数据，检查是否存在与已知后门触发条件匹配的输入。(2)输出监测观察模型结果，判断是否存在异常。(3)监控模型内部的行为，如神经元的激活状态、中间层输出等。

下载: 导出CSV

参考文献(233)

[1]	ZHANG Jun. Aeronautical mobile communication: The evolution from narrowband to broadband[J]. Engineering, 2021, 7(4): 431–434. doi: 10.1016/j.eng.2021.02.002.
[2]	POST J. The next generation air transportation system of the united states: Vision, accomplishments, and future directions[J]. Engineering, 2021, 7(4): 427–430. doi: 10.1016/j.eng.2020.05.026.
[3]	BOLIĆ T and RAVENHILL P. SESAR: The past, present, and future of European air traffic management research[J]. Engineering, 2021, 7(4): 448–451. doi: 10.1016/j.eng.2020.08.023.
[4]	中国民用航空局. 智慧民航建设路线图[R]. 2022. Civil Aviation Administration of China. Roadmap for building intelligent civil aviation[R]. 2022. (查阅网上资料, 未找到本条文献英文和报告编号信息, 请补充) .
[5]	International Civil Aviation Organization. Global air traffic management operational concept[R]. Doc 9854, 2005.
[6]	马兰, 孟诗君, 吴志军. 基于BERT与生成对抗的民航陆空通话意图挖掘[J]. 系统工程与电子技术, 2024, 46(2): 740–750. doi: 10.12305/j.issn.1001-506X.2024.02.38. MA Lan, MENG Shijun, and WU Zhijun. Intention mining for civil aviation radiotelephony communication based on BERT and generative adversarial[J]. Systems Engineering and Electronics, 2024, 46(2): 740–750. doi: 10.12305/j.issn.1001-506X.2024.02.38.
[7]	WU Zhijun, LIANG Cheng, and ZHANG Yuan. Blockchain-based authentication of GNSS civil navigation message[J]. IEEE Transactions on Aerospace and Electronic Systems, 2023, 59(4): 4380–4392. doi: 10.1109/TAES.2023.3241041.
[8]	KHAN H A, KHAN H, GHAFOOR S, et al. A survey on security of automatic dependent surveillance -broadcast (ADS-B) protocol: Challenges, potential solutions and future directions[J]. IEEE Communications Surveys & Tutorials, 2024. (查阅网上资料, 未找到本条文献信息, 请确认) .
[9]	NÖHREN L, SCHAPER M, and TYBURZY L. Towards full ATC automation for aircraft ground movement: A first step[C]. Proceedings of 2024 AIAA DATC/IEEE 43rd Digital Avionics Systems Conference (DASC), San Diego, USA, 2024: 1–9. doi: 10.1109/DASC62030.2024.10749019.
[10]	ENAYATOLLAHI F, ATASHGAH M A A, MALAEK S M B, et al. PBN-based time-optimal terminal air traffic control using cellular automata[J]. IEEE Transactions on Aerospace and Electronic Systems, 2021, 57(3): 1513–1523. doi: 10.1109/TAES.2020.3048787.
[11]	SAMPIGETHAYA K. Aircraft cyber security risk assessment: Bringing air traffic control and cyber-physical security to the forefront[C]. Proceedings of the AIAA Scitech 2019 Forum, San Diego, USA, 2019: 1–17. doi: 10.2514/6.2019-0061.
[12]	BOGODA L, MO J, and BIL C. A systems engineering approach to appraise cybersecurity risks of CNS/ATM and avionics systems[C]. Proceedings of 2019 Integrated Communications, Navigation and Surveillance Conference (ICNS), Herndon, USA, 2019: 1–15. doi: 10.1109/ICNSURV.2019.8735376.
[13]	MEHAN D J. Information systems security: The federal aviation administration’s layered approach[R]. 2000. (查阅网上资料, 未找到本条文献报告编号, 请补充) .
[14]	KIESLING T and KREUZER M. Recommendations to strengthen the cyber resilience of the air traffic system[EB/OL]. http://lb-campus.de/images/content/ARIEL_Recommendations_v2.0.pdf, 2017.
[15]	CSFI. CSFI ATC (Air Traffic Control) cyber security project[EB/OL]. https://csfi.us/pubdocs/?id=47, 2015.
[16]	European Commission. Global ATM security management project[EB/OL]. https://www.gamma-project.eu/, 2024.
[17]	Collaborative Actions for Renovation of Air Traffic Systems. Long-term vision for the future air traffic systems ~ changes to intelligent air traffic systems ~[EB/OL]. https://www.mlit.go.jp/koku/carats/rdk40188/wp-content/uploads/2020/06/LongTermVisionForTheFutureAirTrafficSystemsCARATS.pdf, 2010.
[18]	中国民用航空局. MH/T 0035-2012 民用航空网络与信息安全管理规范[S]. 北京: 中国民用航空局, 2012. Civil Aviation Administration of China. MH/T 0035-2012 Specification for civil aviation network and information security management[S]. Beijing: Civil Aviation Administration of China, 2012. (查阅网上资料, 未找到本条文献出版社信息, 请确认) .
[19]	中国民用航空局. MH/T 0067-2018 民航Web应用系统安全检查指南[S]. 北京: 中国民用航空局, 2018. Civil Aviation Administration of China. MH/T 0067-2018 Security testing guide for Web application system of civil aviation[S]. Beijing: Civil Aviation Administration of China, 2018. (查阅网上资料, 未找到本条文献出版信息, 请确认) .
[20]	中国民用航空局. MH/T 0068-2018 民用航空移动应用程序安全测评指南[S]. 北京: 中国民用航空局, 2018. Civil Aviation Administration of China. MH/T 0068-2018 Security testing guide for mobile application program of civil aviation[S]. Beijing: Civil Aviation Administration of China, 2018. (查阅网上资料, 未找到本条文献出版信息, 请确认) .
[21]	中国民用航空局. MH/T 0069-2018 民用航空网络安全等级保护定级指南[S]. 北京: 中国民用航空局, 2018. Civil Aviation Administration of China. MH/T 0069-2018 Guidelines for grading of classified cyber security protection in civil aviation[S]. Beijing: Civil Aviation Administration of China, 2018. (查阅网上资料, 未找到本条文献出版信息, 请确认) .
[22]	中国民用航空局. MH/T 0076-2020 民用航空网络安全等级保护基本要求[S]. 北京: 中国民用航空局, 2020. Civil Aviation Administration of China. MH/T 0076-2020 Baseline for classified protection of cybersecurity in civil aviation[S]. Beijing: Civil Aviation Administration of China, 2020. (查阅网上资料, 未找到本条文献出版信息, 请确认) .
[23]	PIK E. Airport security: The impact of AI on safety, efficiency, and the passenger experience[J]. Journal of Transportation Security, 2024, 17: 9. doi: 10.1007/S12198-024-00276-6.
[24]	LYKOU G, ANAGNOSTOPOULOU A, and GRITZALIS D. Smart airport cybersecurity: Threat mitigation and cyber resilience controls[J]. Sensors, 2019, 19(1): 19. doi: 10.3390/s19010019.
[25]	LYKOU G, ANAGNOSTOPOULOU A, and GRITZALIS D. Implementing cyber-security measures in airports to improve cyber-resilience[C]. Proceedings of 2018 Global Internet of Things Summit (GIoTS), Bilbao, Spain, 2018: 1–6. doi: 10.1109/GIOTS.2018.8534523.
[26]	KORONIOTIS N, MOUSTAFA N, SCHILIRO F, et al. A holistic review of cybersecurity and reliability perspectives in smart airports[J]. IEEE Access, 2020, 8: 209802–209834. doi: 10.1109/ACCESS.2020.3036728.
[27]	HABLER E, BITTON R, and SHABTAI A. Assessing aircraft security: A comprehensive survey and methodology for evaluation[J]. ACM Computing Surveys, 2024, 56(4): 96. doi: 10.1145/3610772.
[28]	FAKEEH K A. An analysis of airports cyber-security[J]. Communications on Applied Electronics, 2016, 4(7): 11–15. doi: 10.5120/cae2016652129.
[29]	European Union Agency for Network and Information Security. Securing smart airports[R]. 2016. (查阅网上资料, 未找到本条文献报告编号信息, 请补充) .
[30]	DAVE G, CHOUDHARY G, SIHAG V, et al. Cyber security challenges in aviation communication, navigation, and surveillance[J]. Computers & Security, 2022, 112: 102516. doi: 10.1016/j.cose.2021.102516.
[31]	VILLEGAS J, FORTES S, ESCAÑO V, et al. Verification and validation framework for AFDX avionics networks[J]. IEEE Access, 2022, 10: 66743–66756. doi: 10.1109/ACCESS.2022.3184329.
[32]	张双, 孔德岐, 王元勋, 等. 基于虚拟化航电平台的网络域间安全通信技术[J]. 西北工业大学学报, 2022, 40(3): 530–537. doi: 10.1051/jnwpu/20224030530. ZHANG Shuang, KONG Deqi, WANG Yuanxun, et al. Secure communication technology between network domains based on virtualization avionics platform[J]. Journal of Northwestern Polytechnical University, 2022, 40(3): 530–537. doi: 10.1051/jnwpu/20224030530.
[33]	HINTZE H, GIERTZSCH F, and GOD R. Design approach for secure networks to introduce data analytics within the aircraft cabin[J]. SAE International Journal of Advances and Current Practices in Mobility, 2019, 2(2): 737–746. doi: 10.4271/2019-01-1853.
[34]	郭鹏军, 张泾周, 杨远帆, 等. 飞机机内无线通信网络架构与接入控制算法研究[J]. 计算机科学, 2022, 49(9): 268–274. doi: 10.11896/jsjkx.210700220. GUO Pengjun, ZHANG Jingzhou, YANG Yuanfan, et al. Study on wireless communication network architecture and access control algorithm in aircraft[J]. Computer Science, 2022, 49(9): 268–274. doi: 10.11896/jsjkx.210700220.
[35]	WERTHWEIN M, BRUNNER M, and ANNIGHOEFER B. A concept enabling cybersecurity for a self-adaptive avionics platform with respect to RTCA DO-326 and RTCA DO-356[C]. Proceedings of 2023 IEEE/AIAA 42nd Digital Avionics Systems Conference (DASC), Barcelona, Spain, 2023: 1–10. doi: 10.1109/DASC58513.2023.10311289.
[36]	SHAIKH F, RAHOUTI M, GHANI N, et al. A review of recent advances and security challenges in emerging E-enabled aircraft systems[J]. IEEE Access, 2019, 7: 63164–63180. doi: 10.1109/ACCESS.2019.2916617.
[37]	WU Zhijun, YOU Zhenghang, and WANG Peng. Attribute encryption based access control methods under airborne networks[C]. Proceedings of 2022 IEEE Intl Conf on Parallel & Distributed Processing with Applications, Big Data & Cloud Computing, Sustainable Computing & Communications, Social Computing & Networking (ISPA/BDCloud/SocialCom/SustainCom), Melbourne, Australia, 2022: 302–306. doi: 10.1109/ISPA-BDCloud-SocialCom-SustainCom57177.2022.00045.
[38]	PREDESCU A V and STELKENS-KOBSCH T H. Aviation Security Lab: A testbed for security testing of current and future aviation technologies[C]. Proceedings of 2022 IEEE/AIAA 41st Digital Avionics Systems Conference (DASC), Portsmouth, USA, 2022: 1–5. doi: 10.1109/DASC55683.2022.9925750.
[39]	ŠEGVIĆ M, NIKOLIĆ K K, and IVANJKO E. A proposal for a Fully Distributed Flight Control System design[C]. Proceedings of 2016 39th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO), Opatija, Croatia, 2016: 1099–1103. doi: 10.1109/MIPRO.2016.7522304.
[40]	WIELAND F and KILBOURNE T. A digital twin of a flight management system: Findings from the cloud FMS project[C]. Proceedings of the AIAA SCITECH 2024 Forum, Orlando, USA, 2024: 0278. doi: 10.2514/6.2024-0278.
[41]	HUANG Chenyu and CHENG Xiaoyue. Estimation of aircraft fuel consumption by modeling flight data from avionics systems[J]. Journal of Air Transport Management, 2022, 99: 102181. doi: 10.1016/j.jairtraman.2022.102181.
[42]	ZARIKOFF B, MARTIN D, and INSLEY M. Lightweight, low-cost and flexible flight data monitoring[C]. Proceedings of 2014 IEEE AUTOTEST, St. Louis, USA, 2014: 251–259. doi: 10.1109/AUTEST.2014.6935154.
[43]	TANGTHONG N and AKTIMAGOOL S. Management of laboratory-based learning activity on electronic flight instrument system[C]. Proceedings of 2021 6th International STEM Education Conference (iSTEM-Ed), Pattaya, Thailand, 2021: 1–4. doi: 10.1109/iSTEM-Ed52129.2021.9625098.
[44]	KHAN G A I and MAJID I. Modeling and simulation of interference between aircraft radar altimeter and 5G-C band transmissions along with implementation of proposed solution[C]. Proceedings of the International Conference on Aeronautical Sciences, Engineering and Technology, Muscat, Oman, 2024: 176–185. doi: 10.1007/978-981-99-7775-8_18.
[45]	CARROLL M, REBENSKY S, WILT D, et al. Integrating uncertified information from the electronic flight bag into the aircraft panel: Impacts on pilot response[J]. International Journal of Human-Computer Interaction, 2021, 37(7): 630–641. doi: 10.1080/10447318.2020.1854001.
[46]	TURTIAINEN H, COSTIN A, KHANDKER S, et al. GDL90fuzz: Fuzzing-GDL-90 data interface specification within aviation software and avionics devices–a cybersecurity pentesting perspective[J]. IEEE Access, 2022, 10: 21554–21562. doi: 10.1109/ACCESS.2022.3150840.
[47]	MANIKATH E, LI W C, and PIOTROWSKI P. Usability assessment on existing alerting designs for emergency communication between passengers and cabin crews[C]. Proceedings of the German Aerospace Conference (Deutscher Luft- und Raumfahrtkongress) 2023, Stuttgart, Germany, 2024: 610019. doi: 10.25967/610019.
[48]	METWALLY E A and MOHAMMED H T. Hacking an aircraft: Hacking the in-flight entertainment system[J]. Advances in Networks, 2022, 10(1): 7–14. doi: 10.11648/j.net.20221001.12.
[49]	AYUB S, PETRUNIN I, TSOURDOS A, et al. In-flight entertainment datalink analysis and simulation[C]. Proceedings of 2020 AIAA/IEEE 39th Digital Avionics Systems Conference (DASC), San Antonio, USA, 2020: 1–10. doi: 10.1109/DASC50938.2020.9256432.
[50]	苟江, 梁卫星, 田伟, 等. 民机后舱一体化网络架构QoS算法设计[J]. 计算机与网络, 2024, 50(2): 156–162. doi: 10.20149/j.cnki.issn1008-1739.2024.02.011. GOU Jiang, LIANG Weixing, TIAN Wei, et al. Design of QoS algorithm for civil aircraft rear cabin integrated network architecture[J]. Computer & Network, 2024, 50(2): 156–162. doi: 10.20149/j.cnki.issn1008-1739.2024.02.011.
[51]	ZHOU Xuan, HE Feng, ZHAO Luxi, et al. Hybrid scheduling of tasks and messages for TSN-based avionics systems[J]. IEEE Transactions on Industrial Informatics, 2024, 20(2): 1081–1092. doi: 10.1109/TII.2023.3254608.
[52]	赵长啸, 汪克念, 张伟, 等. 民机航电系统功能-信息安全一体化分析方法[J]. 中国安全科学学报, 2022, 32(9): 49–56. doi: 10.16265/j.cnki.issn1003-3033.2022.09.2126. ZHAO Changxiao, WANG Kenian, ZHANG Wei, et al. Integrated analysis method of functional safety and cyber security of avionics system for civil aircraft[J]. China Safety Science Journal, 2022, 32(9): 49–56. doi: 10.16265/j.cnki.issn1003-3033.2022.09.2126.
[53]	SIDDIQUI F, AHLBRECHT A, KHAN R, et al. Cybersecurity engineering: Bridging the security gaps in avionics architectures and DO-326A/ED-202A[C]. Proceedings of 2023 IEEE/AIAA 42nd Digital Avionics Systems Conference (DASC), Barcelona, Spain, 2023: 1–8. doi: 10.1109/DASC58513.2023.10311187.
[54]	RAMANATT P R, NATARAJAN K, and SHOBHA K R. Challenges in implementing a wireless avionics network[J]. Aircraft Engineering and Aerospace Technology, 2020, 92(3): 482–494. doi: 10.1108/AEAT-07-2019-0144.
[55]	KULESHOV Y A, NAGPAL K, UCPINAR K, et al. Cyber attacks on avionics networks in digital twin environment: Detection and defense[C]. Proceedings of the AIAA SCITECH 2024 Forum, Orlando, USA, 2024: 0277. doi: 10.2514/6.2024-0277.
[56]	WU Zhijun, ZHANG Yun, YANG Yiming, et al. Spoofing and anti-spoofing technologies of global navigation satellite system: A survey[J]. IEEE Access, 2020, 8: 165444–165496. doi: 10.1109/ACCESS.2020.3022294.
[57]	VILLAGE A. DEF CON 28 aerospace village: Attacking flight management systems[J]. Retrieved Feb, 2020, 12: 2022. (查阅网上资料, 未找到本条文献信息, 请确认) .
[58]	SHATILIN I. Hacking an aircraft: Is it already real?[R]. 2015. (查阅网上资料, 未找到本条文献报告编号信息, 请补充) .
[59]	ISHTIAQ S and ABD RAHMAN N A. Cybersecurity vulnerabilities and defence techniques in aviation industry[C]. Proceedings of the 3rd International Conference on Integrated Intelligent Computing Communication & Security (ICIIC 2021), 2021: 559–567. (查阅网上资料, 未找到本条文献出版地信息, 请补充) .
[60]	DORIGATTI D, STROHMEIER M, and NEUHAUS S. Air-bus hijacking: Silently taking over avionics systems[C]. Proceedings of the 10th ACM Cyber-Physical System Security Workshop, Singapore, Singapore, 2024: 53–63. doi: 10.1145/3626205.3659144.
[61]	AL-SHAER R, SPRING J M, and CHRISTOU E. Learning the associations of MITRE ATT & CK adversarial techniques[C]. Proceedings of 2020 IEEE Conference on Communications and Network Security (CNS), Avignon, France, 2020: 1–9. doi: 10.1109/CNS48642.2020.9162207.
[62]	HILLEBRECHT A, MARKS T, and GOLLNICK V. An aeronautical data communication demand model for the North Atlantic oceanic airspace[J]. CEAS Aeronautical Journal, 2023, 14(2): 553–567. doi: 10.1007/s13272-023-00651-4.
[63]	BARON C and LOUIS V. Towards a continuous certification of safety-critical avionics software[J]. Computers in Industry, 2021, 125: 103382. doi: 10.1016/j.compind.2020.103382.
[64]	KAMBOJ P, KHARE S, and PAL S. User authentication using Blockchain based smart contract in role-based access control[J]. Peer-to-Peer Networking and Applications, 2021, 14(5): 2961–2976. doi: 10.1007/s12083-021-01150-1.
[65]	KO Y, KIM J, DUGUMA D G, et al. Drone secure communication protocol for future sensitive applications in military zone[J]. Sensors, 2021, 21(6): 2057. doi: 10.3390/s21062057.
[66]	JANGJOU M and SOHRABI M K. A comprehensive survey on security challenges in different network layers in cloud computing[J]. Archives of Computational Methods in Engineering, 2022, 29(6): 3587–3608. doi: 10.1007/s11831-022-09708-9.
[67]	ASANTE M, EPIPHANIOU G, MAPLE C, et al. Distributed ledger technologies in supply chain security management: A comprehensive survey[J]. IEEE Transactions on Engineering Management, 2023, 70(2): 713–739. doi: 10.1109/TEM.2021.3053655.
[68]	CASADO E, RODRIGUEZ R M, TABOSO P, et al. Information security in future air traffic management systems[J]. Journal of Aerospace Information Systems, 2016, 13(3): 101–112. doi: 10.2514/1.I010233.
[69]	PRANDINI M, PIRODDI L, PUECHMOREL S, et al. Toward air traffic complexity assessment in new generation air traffic management systems[J]. IEEE Transactions on Intelligent Transportation Systems, 2011, 12(3): 809–818. doi: 10.1109/TITS.2011.2113175.
[70]	STROHMEIER M, SCHAFER M, PINHEIRO R, et al. On perception and reality in wireless air traffic communication security[J]. IEEE Transactions on Intelligent Transportation Systems, 2017, 18(6): 1338–1357. doi: 10.1109/TITS.2016.2612584.
[71]	杨乐. 民航空管信息系统网络安全态势感知与分析[J]. 管理科学与工程, 2022, 11(4): 665–671. doi: 10.12677/MSE.2022.114079. YANG Le. Network security situation awareness and analysis of civil aviation management information system[J]. Management Science and Engineering, 2022, 11(4): 665–671. doi: 10.12677/MSE.2022.114079.
[72]	CHEN Shichuan, ZHENG Shilian, YANG Lifeng, et al. Deep learning for large-scale real-world ACARS and ADS-B radio signal classification[J]. IEEE Access, 2019, 7: 89256–89264. doi: 10.1109/ACCESS.2019.2925569.
[73]	GULTEPE G, KANAR T, ZIHIR S, et al. A 1024-element Ku-band SATCOM dual-polarized receiver with >10-dB/K G/T and embedded transmit rejection filter[J]. IEEE Transactions on Microwave Theory and Techniques, 2021, 69(7): 3484–3495. doi: 10.1109/TMTT.2021.3073321.
[74]	JIANG Wei, LIU Dan, CAI Baigen, et al. A fault-tolerant tightly coupled GNSS/INS/OVS integration vehicle navigation system based on an FDP algorithm[J]. IEEE Transactions on Vehicular Technology, 2019, 68(7): 6365–6378. doi: 10.1109/TVT.2019.2916852.
[75]	李腾耀, 王布宏, 尚福特, 等. ADS-B 攻击数据弹性恢复方法[J]. 电子与信息学报, 2020, 42(10): 2365–2373. doi: 10.11999/JEIT191020. LI Tengyao, WANG Buhong, SHANG Fute, et al. A resilient recovery method on ADS-B attack data[J]. Journal of Electronics & Information Technology, 2020, 42(10): 2365–2373. doi: 10.11999/JEIT191020.
[76]	罗鹏, 王布宏, 李腾耀. 基于BiGRU-SVDD的ADS-B异常数据检测模型[J]. 航空学报, 2020, 41(10): 323878. doi: 10.7527/S1000-6893.2020.23878. LUO Peng, WANG Buhong, and LI Tengyao. ADS-B anomaly data detection model based on BiGRU-SVDD[J]. Acta Aeronautica et Astronautica Sinica, 2020, 41(10): 323878. doi: 10.7527/S1000-6893.2020.23878.
[77]	董襄宁, 赵征, 张洪海. 空中交通管理基础[M]. 北京: 科学出版社, 2011: 1–3. DONG Xiangning, ZHAO Zheng, and ZHANG Honghai. Fundamentals of Air Traffic Management[M]. Beijing: Science Press, 2011: 1–3. (查阅网上资料, 未找到本条文献英文信息, 请确认) .
[78]	WILLIAMSON T and SPENCER N A. Development and operation of the Traffic Alert and Collision Avoidance System (TCAS)[J]. Proceedings of the IEEE, 1989, 77(11): 1735–1744. doi: 10.1109/5.47735.
[79]	王布宏, 罗鹏, 李腾耀, 等. 基于粒子群优化多核支持向量数据描述的广播式自动相关监视异常数据检测模型[J]. 电子与信息学报, 2020, 42(11): 2727–2734. doi: 10.11999/JEIT190767. WANG Buhong, LUO Peng, LI Tengyao, et al. ADS-B anomalous data detection model based on PSO-MKSVDD[J]. Journal of Electronics & Information Technology, 2020, 42(11): 2727–2734. doi: 10.11999/JEIT190767.
[80]	SHANG Fute, WANG Buhong, YAN Fuhu, et al. Multidevice false data injection attack models of ADS-B multilateration systems[J]. Security and Communication Networks, 2019, 2019(1): 8936784. doi: 10.1155/2019/8936784.
[81]	LI Tengyao, WANG Buhong, SHANG Fute, et al. Threat model and construction strategy on ADS-B attack data[J]. IET Information Security, 2020, 14(5): 542–552. doi: 10.1049/iet-ifs.2018.5635.
[82]	SHANG Fute, WANG Buhong, LI Tengyao, et al. Adversarial examples on deep-learning based ADS-B spoofing detection[J]. IEEE Wireless Communications Letters, 2020, 9(10): 1734–1737. doi: 10.1109/LWC.2020.3002914.
[83]	LI Tengyao, WANG Buhong, SHANG Fute, et al. Dynamic temporal ADS-B data attack detection based on sHDP-HMM[J]. Computers & Security, 2020, 93: 101789. doi: 10.1016/j.cose.2020.101789.
[84]	LI Tengyao, WANG Buhong, SHANG Fute, et al. Online sequential attack detection for ADS-B data based on hierarchical temporal memory[J]. Computers & Security, 2019, 87: 101599. doi: 10.1016/j.cose.2019.101599.
[85]	LI Tengyao and WANG Buhong. Sequential collaborative detection strategy on ADS-B data attack[J]. International Journal of Critical Infrastructure Protection, 2019, 24: 78–99. doi: 10.1016/j.ijcip.2018.11.003.
[86]	LUO Peng, WANG Buhong, LI Tengyao, et al. ADS-B anomaly data detection model based on VAE-SVDD[J]. Computers & Security, 2021, 104: 102213. doi: 10.1016/j.cose.2021.102213. .
[87]	ESKILSSON S, GUSTAFSSON H, KHAN S, et al. Demonstrating ADS-B AND CPDLC attacks with software-defined radio[C]. Proceedings of 2020 Integrated Communications Navigation and Surveillance Conference (ICNS), Herndon, USA, 2020: 1B2-1–1B2-9. doi: 10.1109/ICNS50378.2020.9222945.
[88]	NGUYEN L K, NGUYEN D H N, TRAN N H, et al. SATCOM jamming resiliency under non-uniform probability of attacks[C]. Proceedings of the MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM), San Diego, USA, 2021: 85–90. doi: 10.1109/MILCOM52596.2021.9652944.
[89]	LI Xinwei, ZHANG Qianyun, XU Lexi, et al. A compatible and identity privacy-preserving security protocol for ACARS[C]. Proceedings of 2022 IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom), Wuhan, China, 2022: 1048–1053. doi: 10.1109/TrustCom56396.2022.00143.
[90]	CHOUDHARY G, SIHAG V, GUPTA S, et al. Aviation attacks based on ILS and VOR vulnerabilities[J]. Journal of Surveillance, Security and Safety, 2022, 3: 27–40. doi: 10.20517/jsss.2021.17.
[91]	YANG Zhen, YING Jun, SHEN Junjie, et al. Anomaly detection against GPS spoofing attacks on connected and autonomous vehicles using learning from demonstration[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24(9): 9462–9475. doi: 10.1109/TITS.2023.3269029.
[92]	OBOD I, SVYD I, ZAVOLODKO G, et al. Assessing SSR relative data capacity[C]. Proceedings of 2021 IEEE 3rd Ukraine Conference on Electrical and Computer Engineering (UKRCON), Lviv, Ukraine, 2021: 142–146. doi: 10.1109/UKRCON53503.2021.9575971.
[93]	ZHANG Qianyun, WANG Zhendong, WU Biyi, et al. A Robust and Practical Solution to ADS-B security against denial-of-service attacks[J]. IEEE Internet of Things Journal, 2024, 11(8): 13647–13659. doi: 10.1109/JIOT.2023.3337543.
[94]	MONTEIRO M. Detecting malicious ADS-B broadcasts using wide area multilateration[C]. Proceedings of 2015 IEEE/AIAA 34th Digital Avionics Systems Conference (DASC), Prague, Czech Republic, 2015: 1–28. doi: 10.1109/DASC.2015.7311579.
[95]	LUO Peng, WANG Buhong, and TIAN Jiwei. TTSAD: TCN-Transformer-SVDD Model for Anomaly Detection in air traffic ADS-B data[J]. Computers & Security, 2024, 141: 103840. doi: 10.1016/j.cose.2024.103840.
[96]	LUO Peng, WANG Buhong, TIAN Jiwei, et al. ADS-Bpois: Poisoning attacks against deep-learning-based air traffic ADS-B unsupervised anomaly detection models[J]. IEEE Internet of Things Journal, 2024, 11(23): 38301–38311. doi: 10.1109/JIOT.2024.3446675.
[97]	LUO Peng, WANG Buhong, TIAN Jiwei, et al. Adversarial attacks against deep-learning-based automatic dependent surveillance-broadcast unsupervised anomaly detection models in the context of air traffic management[J]. Sensors, 2024, 24(11): 3584. doi: 10.3390/s24113584.
[98]	MCPARLAND T, PATEL V, and HUGHES W J. Securing air-ground communications[C]. Proceedings of the 20th Digital Avionics Systems Conference, Daytona Beach, USA, 2001: 7A7/1–7A7/9. doi: 10.1109/DASC.2001.964187.
[99]	SCHRAML M G and KNOPP A. Physical layer security with unknown eavesdroppers in beyond-5G MU-MIMO SATCOM[C]. Proceedings of 2020 IEEE 3rd 5G World Forum (5GWF), Bangalore, India, 2020: 180–185. doi: 10.1109/5GWF49715.2020.9221107.
[100]	YUE Meng and WU Xiaofeng. The approach of ACARS data encryption and authentication[C]. Proceedings of 2010 International Conference on Computational Intelligence and Security, Nanning, China, 2010: 556–560. doi: 10.1109/CIS.2010.127.
[101]	ADAMY D. EW 102: A Second Course in Electronic Warfare[M]. Norwood, USA: Artech House, 2003. (查阅网上资料, 未找到本条文献出版年和页码信息, 请确认并补充) .
[102]	BLANCH J, WALTER T, and ENGE P. Satellite navigation for aviation in 2025[J]. Proceedings of the IEEE, 2012, 100(Special Centennial Issue): 1821–1830. doi: 10.1109/JPROC.2012.2190154.
[103]	STROHMEIER M, LENDERS V, and MARTINOVIC I. Intrusion detection for airborne communication using PHY-layer information[C]. Proceedings of the 12th International Conference on Detection of Intrusions and Malware, and Vulnerability Assessment, Milan, Italy, 2015: 67–77. doi: 10.1007/978-3-319-20550-2_4.
[104]	CHIANG J T, HAAS J J, CHOI J, et al. Secure location verification using simultaneous multilateration[J]. IEEE Transactions on Wireless Communications, 2012, 11(2): 584–591. doi: 10.1109/TWC.2011.120911.101147.
[105]	MUNIR A, AVED A, and BLASCH E. Situational awareness: Techniques, challenges, and prospects[J]. AI, 2022, 3(1): 55–77. doi: 10.3390/ai3010005.
[106]	WU Zhijun, XU Pei, and FAN Haoyu. Network security situation assessment method based eigenvector centrality[C]. Proceedings of 2024 International Wireless Communications and Mobile Computing (IWCMC), Ayia Napa, Cyprus, 2024: 103–108. doi: 10.1109/IWCMC61514.2024.10592357.
[107]	WU Zhijun and FAN Haoyu. Aviation network security situation awareness based on game theory[C]. Proceedings of 2023 IEEE Aerospace Conference, Big Sky, USA, 2023: 1–8. doi: 10.1109/AERO55745.2023.10115582.
[108]	WU Zhijun, BAI Zhuoning, ZHANG Lizhe, et al. Feature extraction method based on sparse autoencoder for air traffic management system security situation awareness[J]. Security and Communication Networks, 2022, 2022(1): 3757662. doi: 10.1155/2022/3757662.
[109]	MA Lan, MA Shaopu, and WU Zhijun. WNN-based prediction of security situation awareness for the civil aviation network[J]. Journal of Intelligent Systems, 2014, 24(1): 55–67. doi: 10.1515/jisys-2014-0004.
[110]	SMITH M, MOSER D, STROHMEIER M, et al. Undermining privacy in the Aircraft Communications Addressing and Reporting System (ACARS)[J]. Proceedings on Privacy Enhancing Technologies, 2018, 2018(3): 105–122. doi: 10.1515/popets-2018-0023.
[111]	UKWANDU E, BEN-FARAH M, HINDY H, et al. Cyber-security challenges in aviation industry: A review of current and future trends[J]. Information, 2022, 13(3): 146. doi: 10.3390/info13030146.
[112]	吴志军. 广域信息管理SWIM信息安全关键技术[M]. 北京: 人民邮电出版社, 2020: 13–19. WU Zhijun. Key Technologies of Information Assurance for System Wide Information Management (SWIM)[M]. Beijing: Posts & Telecommunications Press, 2020: 13–19.
[113]	李晛. 基于物联网技术的智慧机场设计与应用探讨[J]. 工程管理, 2024, 5(7): 14–15. doi: 10.12238/jpm.v5i7.6950. LI Xian. Discussion on the smart airport design and application based on the Internet of Things technology[J]. Journal of Project Management, 2024, 5(7): 14–15. doi: 10.12238/jpm.v5i7.6950. (查阅网上资料,未找到本条文献英文信息,请确认) .
[114]	顾兆军, 张一诺, 宋跃东, 等. 智慧机场物联网应用及网络安全挑战[J]. 指挥信息系统与技术, 2023, 14(5): 14–20. doi: 10.15908/j.cnki.cist.2023.05.003. GU Zhaojun, ZHANG Yinuo, SONG Yuedong, et al. IoT applications and cyber security challenges in smart airports[J]. Command Information System and Technology, 2023, 14(5): 14–20. doi: 10.15908/j.cnki.cist.2023.05.003.
[115]	MAJID S A, NUGRAHA A, SULISTIYONO B S, et al. The effect of safety risk management and airport personnel competency on aviation safety performance[J]. Uncertain Supply Chain Management, 2022, 10(4): 1509–1522. doi: 10.5267/j.uscm.2022.6.004.
[116]	XIONG Wenjun and LAGERSTRÖM R. Threat modeling–a systematic literature review[J]. Computers & Security, 2019, 84: 53–69. doi: 10.1016/j.cose.2019.03.010.
[117]	邢馨心, 左青雅, 刘建伟. 基于5G的智慧机场网络安全方案设计与安全性分析[J]. 网络与信息安全学报, 2023, 9(5): 116–126. doi: 10.11959/j.issn.2096-109x.2023075. XING Xinxin, ZUO Qingya, and LIU Jianwei. 5G-based smart airport network security scheme design and security analysis[J]. Chinese Journal of Network and Information Security, 2023, 9(5): 116–126. doi: 10.11959/j.issn.2096-109x.2023075.
[118]	FLORIDO-BENÍTEZ L. Identifying and classifying cyberattacks on airports[J]. Cyber Security: A Peer-Reviewed Journal, 2024, 8(1): 63–79. doi: 10.69554/TWHW5595.
[119]	SUCIU G, SCHEIANU A, VULPE A, et al. Cyber-attacks-the impact over airports security and prevention modalities[C]. Proceedings of the WorldCIST: World Conference on Information Systems and Technologies, Naples, Italy, 2018: 154–162. doi: 10.1007/978-3-319-77700-9_16. (查阅网上资料,未找到本条文献出版地信息,请确认) .
[120]	KALAIVANI N, RAMAN R, BASAVARADDI C C S, et al. Enhancing air travel with IoT: Smart airports and passenger experience[C]. Proceedings of 2023 7th International Conference on Electronics, Communication and Aerospace Technology (ICECA), Coimbatore, India, 2023: 1300–1305. doi: 10.1109/ICECA58529.2023.10394746.
[121]	SAADA H, ORIZIO R, and SEBASTIO S. Modeling and conducting security risk assessment of smart airport infrastructures with SecRAM[C]. Proceedings of the 7th International Conference on Networking, Intelligent Systems and Security, Meknes, Morocco, 2024: 59. doi: 10.1145/3659677.3659992.
[122]	ALLABERGANOV B A and ABDULLAYEV D. Models for protecting airport information systems from cyber incidents[J]. Science and Innovation, 2022, 1(8): 919–923. doi: 10.5281/zenodo.7445057.
[123]	FERON E M, SANNI O, MOTE M, et al. Ariadne: A common-sense thread for enabling provable safety in air mobility systems with unreliable components[C]. Proceedings of the AIAA SciTech 2022 Forum, San Diego, USA, 2022: 0057. doi: 10.2514/6.2022-0057.
[124]	ABOAOJA F A, ZAINAL A, GHALEB F A, et al. Malware detection issues, challenges, and future directions: A survey[J]. Applied Sciences, 2022, 12(17): 8482. doi: 10.3390/app12178482.
[125]	ABELSON H, ANDERSON R, BELLOVIN S M, et al. Bugs in our pockets: The risks of client-side scanning[J]. Journal of Cybersecurity, 2024, 10(1): tyad020. doi: 10.1093/cybsec/tyad020.
[126]	LI Tong, WANG Xiaowei, and NI Yeming. Aligning social concerns with information system security: A fundamental ontology for social engineering[J]. Information Systems, 2022, 104: 101699. doi: 10.1016/j.is.2020.101699.
[127]	ELSALLAMY S and ABD RAHMAN N A. AI implementation in airport system: A study of vulnerabilities and countermeasures[J]. Journal of Applied Technology and Innovation, 2022, 6(2): 11–16.
[128]	SEN S, AYDOGAN E, and AYSAN A I. Coevolution of mobile malware and anti-malware[J]. IEEE Transactions on Information Forensics and Security, 2018, 13(10): 2563–2574. doi: 10.1109/TIFS.2018.2824250.
[129]	丁磊, 韩建云, 张西武, 等. 智慧机场物联网系统安全防护研究[J]. 民航学报, 2021, 5(5): 81–84. doi: 10.3969/j.issn.2096-4994.2021.05.021. DING Lei, HAN Jianyun, ZHANG Xiwu, et al. Research on cyber security protection of IOT system of smart airport[J]. Journal of Civil Aviation, 2021, 5(5): 81–84. doi: 10.3969/j.issn.2096-4994.2021.05.021.
[130]	SCOZZARO G, MUJICA MOTA M, DELAHAYE D, et al. Simulation-optimisation-based decision support system for managing airport security resources[C]. Proceedings of the 11th Congress on Simulation for a Sustainable Future, Amsterdam, The Netherlands, 2023: 140–155. doi: 10.1007/978-3-031-68438-8_11.
[131]	ALSULAMI A A and ZEIN-SABATTO S. Resilient cyber-security approach for aviation cyber-physical systems protection against sensor spoofing attacks[C]. Proceedings of 2021 IEEE 11th Annual Computing and Communication Workshop and Conference (CCWC), USA, 2021: 565–571. doi: 10.1109/CCWC51732.2021.9376158. (查阅网上资料,未找到本条文献出版地信息,请补充) .
[132]	ABDELGHANI T. Implementation of defense in depth strategy to secure industrial control system in critical infrastructures[J]. American Journal of Artificial Intelligence, 2019, 3(2): 17–22. doi: 10.11648/j.ajai.20190302.11.
[133]	VYAS S, HANNAY J, BOLTON A, et al. Automated cyber defence: A review[EB/OL]. https://arxiv.org/abs/2303.04926, 2023.
[134]	ALABBAD M, MHASKAR N, and KHEDRI R. Hardening of network segmentation using automated referential penetration testing[J]. Journal of Network and Computer Applications, 2024, 224: 103851. doi: 10.1016/j.jnca.2024.103851.
[135]	ASHRAFI R and ALKINDI H. A framework for IS/IT disaster recovery planning[J]. International Journal of Business Continuity and Risk Management, 2022, 12(1): 1–21. doi: 10.1504/IJBCRM.2022.121645.
[136]	BEAN B. Mitigating insider threats in the domestic aviation system: Policy options for TSA[J]. Homeland Security Affairs, 2017. (查阅网上资料, 未找到本条文献信息, 请确认) .
[137]	CORNACCHIA M, PAPA F, and SAPIO B. User acceptance of voice biometrics in managing the physical access to a secure area of an international airport[J]. Technology Analysis & Strategic Management, 2020, 32(10): 1236–1250. doi: 10.1080/09537325.2020.1758655.
[138]	CURRAH P and MULQUEEN T. Securitizing gender: Identity, biometrics, and transgender bodies at the airport[J]. Social Research: An International Quarterly, 2011, 78(2): 557–582. doi: 10.1353/sor.2011.0030.
[139]	KARAMITSOS I, PAPADAKI M, AL-HUSSAENI K, et al. Transforming airport security: Enhancing efficiency through blockchain smart contracts[J]. Electronics, 2023, 12(21): 4492. doi: 10.3390/electronics12214492.
[140]	MALIK H, TAHIR S, TAHIR H, et al. A homomorphic approach for security and privacy preservation of Smart Airports[J]. Future Generation Computer Systems, 2023, 141: 500–513. doi: 10.1016/j.future.2022.12.005.
[141]	ALI A S and HASAN D S. An IoT-based smart airport check-in system via three-factor authentication (3FA)[J]. Zanco Journal of Pure and Applied Sciences, 2023, 35(4): 1–13. doi: 10.21271/ZJPAS.35.4.01.
[142]	NIRAULA M. Cybersecurity and interoperability of aviation safety service ecosystem[C]. Proceedings of 2022 Integrated Communication, Navigation and Surveillance Conference (ICNS), Dulles, USA, 2022: 1–12. doi: 10.1109/ICNS54818.2022.9771482.
[143]	SUCIU G, SCHEIANU A, PETRE I, et al. Cybersecurity threats analysis for airports[C]. Proceedings of the World Conference on Information Systems and Technologies, Galicia, Spain, 2019: 252–262. doi: 10.1007/978-3-030-16184-2_25.
[144]	ZORLU O, OZSOY A, SERT S A. A role-based access control management model on blockchain for restricted facilities: An airport example[C]. Proceedings of 2023 10th International Conference on Recent Advances in Air and Space Technologies (RAST), Istanbul, Turkiye, 2023: 1–6. doi: 10.1109/RAST57548.2023.10197974.
[145]	WILLEMSEN B and CADEE M. Extending the airport boundary: Connecting physical security and cybersecurity[J]. Journal of Airport Management, 2018, 12(3): 236–247. doi: 10.69554/UVLU6436.
[146]	MENZEL D and HESTERMAN J. Airport security threats and strategic options for mitigation[J]. Journal of Airport Management, 2018, 12(2): 118–131. doi: 10.69554/EFBM2046.
[147]	ZHOU Zhipeng, YU Xinhui, ZHU Zeyu, et al. Development and application of a Bayesian network-based model for systematically reducing safety risks in the commercial air transportation system[J]. Safety Science, 2023, 157: 105942. doi: 10.1016/j.ssci.2022.105942.
[148]	EILSTRUP-SANGIOVANNI M. Ordering global governance complexes: The evolution of the governance complex for international civil aviation[J]. The Review of International Organizations, 2022, 17(2): 293–322. doi: 10.1007/s11558-020-09411-z.
[149]	Federal Aviation Administration. NAS operational view[EB/OL]. https://www.faa.gov/about/office_org/headquarters_offices/ang/offices/tc/library/Storyboard/nextgen-overview.html, 2022.
[150]	SUN Mengyuan, TIAN Yong, WANG Xunuo, et al. Transport causality knowledge-guided GCN for propagated delay prediction in airport delay propagation networks[J]. Expert Systems with Applications, 2024, 240: 122426. doi: 10.1016/j.eswa.2023.122426.
[151]	CAI Kaiquan, LI Yue, ZHU Yongwen, et al. A geographical and operational deep graph convolutional approach for flight delay prediction[J]. Chinese Journal of Aeronautics, 2023, 36(3): 357–367. doi: 10.1016/j.cja.2022.10.004.
[152]	CAI Kaiquan, LI Yue, FANG Yiping, et al. A deep learning approach for flight delay prediction through time-evolving graphs[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(8): 11397–11407. doi: 10.1109/TITS.2021.3103502.
[153]	GUO Ziyu, MEI Guangxu, LIU Shijun, et al. SGDAN—A spatio-temporal graph dual-attention neural network for quantified flight delay prediction[J]. Sensors, 2020, 20(22): 6433. doi: 10.3390/s20226433.
[154]	LI Chi, QI Xixian, YANG Yuzhe, et al. FAST-CA: Fusion-based adaptive spatial-temporal learning with coupled attention for airport network delay propagation prediction[J]. Information Fusion, 2024, 107: 102326. doi: 10.1016/j.inffus.2024.102326.
[155]	RUAN J H, WANG Z X, CHAN F T S, et al. A reinforcement learning-based algorithm for the aircraft maintenance routing problem[J]. Expert Systems with Applications, 2021, 169: 114399. doi: 10.1016/j.eswa.2020.114399.
[156]	OSZUST M, KAPUSCINSKI T, WARCHOL D, et al. A vision-based method for supporting autonomous aircraft landing[J]. Aircraft Engineering and Aerospace Technology, 2018, 90(6): 973–982. doi: 10.1108/AEAT-11-2017-0250.
[157]	GAO Ruizhen, CHEN Meng, ZHAO Ziyue, et al. Real-time detection algorithm of aircraft landing gear based on improved YOLOv8[J]. 2024. doi: 10.21203/rs.3.rs-4493909/v1. (查阅网上资料,未找到本条文献刊名卷期号和页码信息,请补充) .
[158]	AMIT R A and MOHAN C K. A robust airport runway detection network based on R-CNN using remote sensing images[J]. IEEE Aerospace and Electronic Systems Magazine, 2021, 36(11): 4–20. doi: 10.1109/MAES.2021.3088477.
[159]	LIU Mingkun, FENG Guangkun, XU Tingbing, et al. Fusing dense features and pose consistency: A regression method for attitude measurement of aircraft landing[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 5007913. doi: 10.1109/TIM.2023.3244803.
[160]	SHI Qingbang and LI Jun. Objects detection of UAV for anti-UAV based on YOLOv4[C]. Proceedings of 2020 IEEE 2nd International Conference on Civil Aviation Safety and Information Technology, Weihai, China, 2020: 1048–1052. doi: 10.1109/ICCASIT50869.2020.9368788.
[161]	ZHAO Jie, ZHANG Jingshu, LI Dongdong, et al. Vision-based anti-UAV detection and tracking[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(12): 25323–25334. doi: 10.1109/TITS.2022.3177627.
[162]	WANG Chuanyu, WANG Tian, WANG Ershen, et al. Flying small target detection for anti-UAV based on a Gaussian mixture model in a compressive sensing domain[J]. Sensors, 2019, 19(9): 2168. doi: 10.3390/s19092168.
[163]	KANAVOS A, KOUNELIS F, ILIADIS L, et al. Deep learning models for forecasting aviation demand time series[J]. Neural Computing and Applications, 2021, 33(23): 16329–16343. doi: 10.1007/s00521-021-06232-y.
[164]	GU Weifan, GUO Baohua, ZHANG Zhezhe, et al. Civil aviation passenger traffic forecasting: Application and comparative study of the seasonal autoregressive integrated moving average model and backpropagation neural network[J]. Sustainability, 2024, 16(10): 4110. doi: 10.3390/su16104110.
[165]	BRITTAIN M and WEI Peng. Scalable autonomous separation assurance with heterogeneous multi-agent reinforcement learning[J]. IEEE Transactions on Automation Science and Engineering, 2022, 19(4): 2837–2848. doi: 10.1109/TASE.2022.315160.
[166]	KRAVARIS T, LENTZOS K, SANTIPANTAKIS G, et al. Explaining deep reinforcement learning decisions in complex multiagent settings: Towards enabling automation in air traffic flow management[J]. Applied Intelligence, 2023, 53(4): 4063–4098. doi: 10.1007/s10489-022-03605-1.
[167]	WANG Yuan, CAI Weilin, TU Yilei, et al. Reinforcement-learning-informed prescriptive analytics for air traffic flow management[J]. IEEE Transactions on Automation Science and Engineering, 2024, 21(3): 4188–4202. doi: 10.1109/TASE.2023.3292921.
[168]	SPATHARIS C, BASTAS A, KRAVARIS T, et al. Hierarchical multiagent reinforcement learning schemes for air traffic management[J]. Neural Computing and Applications, 2023, 35(1): 147–159. doi: 10.1007/S00521-021-05748-7.
[169]	PAPADOPOULOS G, BASTAS A, VOUROS G A, et al. Deep reinforcement learning in service of air traffic controllers to resolve tactical conflicts[J]. Expert Systems with Applications, 2024, 236: 121234. doi: 10.1016/j.eswa.2023.121234.
[170]	BAUMANN S and KLINGAUF U. Modeling of aircraft fuel consumption using machine learning algorithms[J]. CEAS Aeronautical Journal, 2020, 11(1): 277–287. doi: 10.1007/s13272-019-00422-0.
[171]	ZHU Xinting and LI Lishuai. Flight time prediction for fuel loading decisions with a deep learning approach[J]. Transportation Research Part C: Emerging Technologies, 2021, 128: 103179. doi: 10.1016/j.trc.2021.103179.
[172]	METLEK S. A new proposal for the prediction of an aircraft engine fuel consumption: A novel CNN-BiLSTM deep neural network model[J]. Aircraft Engineering and Aerospace Technology, 2023, 95(5): 838–848. doi: 10.1108/AEAT-05-2022-0132.
[173]	EUROCONTROL. AI/ML based augmented 4D trajectory[EB/OL]. https://www.eurocontrol.int/publication/aiml-based-augmented-4d-trajectory, 2021.
[174]	SHI Zhiyuan, XU Min, PAN Quan, et al. LSTM-based flight trajectory prediction[C]. Proceedings of 2018 International Joint Conference on Neural Networks (IJCNN), Rio de Janeiro, Brazil, 2018: 1–8. doi: 10.1109/IJCNN.2018.8489734.
[175]	ZHANG Xiaoge and MAHADEVAN S. Bayesian neural networks for flight trajectory prediction and safety assessment[J]. Decision Support Systems, 2020, 131: 113246. doi: 10.1016/j.dss.2020.113246.
[176]	ZHANG Zheng, GUO Dongyue, ZHOU Shizhong, et al. Flight trajectory prediction enabled by time-frequency wavelet transform[J]. Nature Communications, 2023, 14(1): 5258. doi: 10.1038/s41467-023-40903-9.
[177]	SHI Zhiyuan, XU Min, and PAN Quan. 4-D flight trajectory prediction with constrained LSTM network[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(11): 7242–7255. doi: 10.1109/TITS.2020.3004807.
[178]	CHEN C J, HUANG C N, and YANG S M. Application of deep learning to multivariate aviation weather forecasting by long short-term memory[J]. Journal of Intelligent & Fuzzy Systems, 2023, 44(3): 4987–4997. doi: 10.3233/JIFS-223183.
[179]	YUAN Ligang, ZENG Yang, CHEN Haiyan, et al. Terminal traffic situation prediction model under the influence of weather based on deep learning approaches[J]. Aerospace, 2022, 9(10): 580. doi: 10.3390/aerospace9100580.
[180]	CHEN C J, HUANG C N, and YANG S M. Aviation visibility forecasting by integrating Convolutional Neural Network and long short-term memory network[J]. Journal of Intelligent & Fuzzy Systems, 2023, 45(3): 5007–5020. doi: 10.3233/JIFS-230483.
[181]	ZHAO Tianyu, ESCRIBANO J, MAJUMDAR A, et al. Spatiotemporal thunderstorm forecasting for pre-tactical air traffic operation: A deep learning approach[C]. Proceedings of AIAA Aviation Forum and Ascend 2024, Las Vegas, Nevada, 2024: 4634. doi: 10.2514/6.2024-4634.
[182]	AYHAN B, VARGO E P, and TANG Huang. On the exploration temporal fusion transformers for anomaly detection with multivariate aviation time-series data[J]. Aerospace, 2024, 11(8): 646. doi: 10.3390/aerospace11080646.
[183]	ZHOU Wentao, CAI Chengtao, WU Kejun, et al. Las-yolo: A lightweight detection method based on YOLOv7 for small objects in airport surveillance[J]. The Journal of Supercomputing, 2024, 80(15): 21764–21789. doi: 10.1007/S11227-024-06289-1.
[184]	ZHANG Xiang, SHU Chang, LI Shuai, et al. AGVS: A new change detection dataset for airport ground video surveillance[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(11): 20588–20600. doi: 10.1109/TITS.2022.3184978.
[185]	LI Weidong, LIU Jia, and MEI Hang. Lightweight convolutional neural network for aircraft small target real-time detection in Airport videos in complex scenes[J]. Scientific Reports, 2022, 12(1): 14474. doi: 10.1038/s41598-022-18263-z.
[186]	ZHOU Wentao, CAI Chengtao, ZHENG Liying, et al. ASSD-YOLO: A small object detection method based on improved YOLOv7 for airport surface surveillance[J]. Multimedia Tools and Applications, 2024, 83(18): 55527–55548. doi: 10.1007/S11042-023-17628-4.
[187]	ZHANG Mengmei, WANG Xiao, ZHU Meiqi, et al. Robust heterogeneous graph neural networks against adversarial attacks[C]. Proceedings of the 36th AAAI Conference on Artificial Intelligence, 2022, 36(4): 4363–4370. doi: 10.1609/aaai.v36i4.20357.
[188]	LIN Xixun, ZHOU Chuan, WU Jia, et al. Exploratory adversarial attacks on graph neural networks for semi-supervised node classification[J]. Pattern Recognition, 2023, 133: 109042. doi: 10.1016/j.patcog.2022.109042.
[189]	FINKELSHTEIN B, BASKIN C, ZHELTONOZHSKII E, et al. Single-node attacks for fooling graph neural networks[J]. Neurocomputing, 2022, 513: 1–12. doi: 10.1016/j.neucom.2022.09.115.
[190]	ZHANG Haotian and MA Xu. Misleading attention and classification: An adversarial attack to fool object detection models in the real world[J]. Computers & Security, 2022, 122: 102876. doi: 10.1016/j.cose.2022.102876.
[191]	LI Guopeng, XU Yue, DING Jian, et al. Towards generic and controllable attacks against object detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 5635812. doi: 10.1109/TGRS.2024.3417958.
[192]	WU Tao, WANG Xuechun, QIAO Shaojie, et al. Small perturbations are enough: Adversarial attacks on time series prediction[J]. Information Sciences, 2022, 587: 794–812. doi: 10.1016/j.ins.2021.11.007.
[193]	KARIM F, MAJUMDAR S, and DARABI H. Adversarial attacks on time series[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(10): 3309–3320. doi: 10.1109/TPAMI.2020.2986319.
[194]	LIU Guanlin and LAI Lifeng. Efficient adversarial attacks on online multi-agent reinforcement learning[C]. Proceedings of the 37th International Conference on Neural Information Processing Systems, New Orleans, USA, 2023: 1062.
[195]	YOU Qiaoben, YING Chengyang, ZHOU Xinning, et al. Understanding adversarial attacks on observations in deep reinforcement learning[J]. Science China Information Sciences, 2024, 67(5): 152104. doi: 10.1007/s11432-021-3688-y.
[196]	SHEN Yun, HE Xinlei, HAN Yufei, et al. Model stealing attacks against inductive graph neural networks[C]. Proceedings of 2022 IEEE Symposium on Security and Privacy (SP), San Francisco, USA, 2022: 1175–1192. doi: 10.1109/SP46214.2022.9833607.
[197]	HE X, JIA J, BACKES M, et al. Stealing links from graph neural networks[C/OL]. Proceedings of the 30th USENIX Security Symposium (USENIX Security 21), 2021: 2669–2686. (查阅网上资料, 未找到本条文献信息, 请确认) .
[198]	PODHAJSKI M, DUBIŃSKI J, BOENISCH F, et al. Efficient model-stealing attacks against inductive graph neural networks[J]. Frontiers in Artificial Intelligence and Applications, 2024, 392: 1438–1445. doi: 10.3233/FAIA240646.
[199]	XU Jing, XUE Minhui, and PICEK S. Explainability-based backdoor attacks against graph neural networks[C]. Proceedings of the 3rd ACM Workshop on Wireless Security and Machine Learning, Abu Dhabi, United Arab Emirates, 2021: 31–36. doi: 10.1145/3468218.3469046.
[200]	ZHENG Haibin, XIONG Haiyang, CHEN Jinyin, et al. Motif-backdoor: Rethinking the backdoor attack on graph neural networks via motifs[J]. IEEE Transactions on Computational Social Systems, 2024, 11(2): 2479–2493. doi: 10.1109/TCSS.2023.3267094.
[201]	CHAN S H, DONG Yinpeng, ZHU Jun, et al. BadDet: Backdoor attacks on object detection[C]. Proceedings of the European Conference on Computer Vision, Tel Aviv, Israel, 2022: 396–412. doi: 10.1007/978-3-031-25056-9_26.
[202]	LUO Chengxiao, LI Yiming, JIANG Yong, et al. Untargeted backdoor attack against object detection[C]. Proceedings of 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Rhodes Island, Greece, 2023: 1–5. doi: 10.1109/ICASSP49357.2023.10095980.
[203]	KIOURTI P, WARDEGA K, JHA S, et al. TrojDRL: Evaluation of backdoor attacks on deep reinforcement learning[C]. Proceedings of 2020 57th ACM/IEEE Design Automation Conference (DAC), San Francisco, USA, 2020: 1–6. doi: 10.1109/DAC18072.2020.9218663.
[204]	CHEN Yanjiao, ZHENG Zhicong, and GONG Xueluan. MARNet: Backdoor attacks against cooperative multi-agent reinforcement learning[J]. IEEE Transactions on Dependable and Secure Computing, 2023, 20(5): 4188–4198. doi: 10.1109/TDSC.2022.3207429.
[205]	HASHEMI S M, BOTEZ R M, and GRIGORIE T L. New reliability studies of data-driven aircraft trajectory prediction[J]. Aerospace, 2020, 7(10): 145. doi: 10.3390/aerospace7100145.
[206]	DONG Guimin, TANG Mingyue, WANG Zhiyuan, et al. Graph neural networks in IoT: A survey[J]. ACM Transactions on Sensor Networks, 2023, 19(2): 47. doi: 10.1145/3565973.
[207]	KIM S and PARK K J. A survey on machine-learning based security design for cyber-physical systems[J]. Applied Sciences, 2021, 11(12): 5458. doi: 10.3390/app11125458.
[208]	VAN IERSEL Q G, MURRIETA MENDOZA A, FELIX PATRON R S, et al. Attack and defense on aircraft trajectory prediction algorithms[C]. Proceedings of the AIAA AVIATION 2022 Forum, Chicago, USA, 2022: 4027. doi: 10.2514/6.2022-4027.
[209]	KIANPOUR M and WEN Shaofang. Timing attacks on machine learning: State of the art[C]. Proceedings of 2019 Intelligent Systems Conference, London, UK, 2020: 111–125. doi: 10.1007/978-3-030-29516-5_10.
[210]	MADRY A, MAKELOV A, SCHMIDT L, et al. Towards deep learning models resistant to adversarial attacks[C]. Proceedings of the 6th International Conference on Learning Representations, Vancouver, Canada, 2017: 1–23. (查阅网上资料, 不确定本条文献标黄作者信息, 请确认) .
[211]	MIYATO T, MAEDA S, KOYAMA M, et al. Virtual adversarial training: A regularization method for supervised and semi-supervised learning[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 41(8): 1979–1993. doi: 10.1109/TPAMI.2018.2858821.
[212]	YE Hongwei, LIU Xiaozhang, and LI Chunlai. DSCAE: A denoising sparse convolutional autoencoder defense against adversarial examples[J]. Journal of Ambient Intelligence and Humanized Computing, 2022, 13(3): 1419–1429. doi: 10.1007/s12652-020-02642-3.
[213]	XU Weilin, EVANS D, and QI Yanjun. Feature squeezing: Detecting adversarial examples in deep neural networks[EB/OL]. https://arxiv.org/abs/1704.01155, 2017.
[214]	TRAMÈR F, KURAKIN A, PAPERNOT N, et al. Ensemble adversarial training: Attacks and defenses[EB/OL]. https://arxiv.org/abs/1705.07204, 2017.
[215]	KUZLU M, CATAK F O, CALI U, et al. Adversarial security mitigations of mmWave beamforming prediction models using defensive distillation and adversarial retraining[J]. International Journal of Information Security, 2023, 22(2): 319–332. doi: 10.1007/s10207-022-00644-0.
[216]	TSENG F H, ZENG Jiangyi, CHO H H, et al. Detecting adversarial examples of fake news via the neurons activation state[J]. IEEE Transactions on Computational Social Systems, 2024, 11(4): 5199–5209. doi: 10.1109/TCSS.2023.3293718.
[217]	ZHANG Lei, ZHOU Yuhang, YANG Yi, et al. Meta invariance defense towards generalizable robustness to unknown adversarial attacks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024, 46(10): 6669–6687. doi: 10.1109/TPAMI.2024.3385745.
[218]	KHALED K, DHAOUADI M, DE MAGALHAES F G, et al. Efficient defense against model stealing attacks on convolutional neural networks[C]. 2023 International Conference on Machine Learning and Applications (ICMLA), Jacksonville, USA, 2023: 45–52. DOI: 10.1109/ICMLA58977.2023.00015.
[219]	郭鑫. 基于决策空间的模型窃取攻击检测与防御方法[D]. [硕士论文], 西安电子科技大学, 2023. GUO Xin. Decision space based approach to detect and defend against model stealing attacks[D]. [Master dissertation], Xidian University, 2023.
[220]	LI Yiming, ZHU Linghui, JIA Xiaojun, et al. Defending against model stealing via verifying embedded external features[C]. Proceedings of the 36th AAAI Conference on Artificial Intelligence, Palo Alto, USA, 2022: 1464–1472. doi: 10.1609/aaai.v36i2.20036. (查阅网上资料,未找到本条文献出版地信息,请确认) .
[221]	THUDUMU S, BRANCH P, JIN Jiong, et al. A comprehensive survey of anomaly detection techniques for high dimensional big data[J]. Journal of Big Data, 2020, 7: 42. doi: 10.1186/s40537-020-00320-x.
[222]	GU Tianyu, DOLAN-GAVITT B, and GARG S. BadNets: Identifying vulnerabilities in the machine learning model supply chain[EB/OL]. https://arxiv.org/abs/1708.06733, 2017.
[223]	WAN Li, ZEILER M, ZHANG Sixin, et al. Regularization of neural networks using dropconnect[C]. Proceedings of the 30th International Conference on International Conference on Machine Learning, Atlanta, USA, 2013: III-1058–III-1066.
[224]	QIAN Zhuang, HUANG Kaizhu, WANG Qiufeng, et al. A survey of robust adversarial training in pattern recognition: Fundamental, theory, and methodologies[J]. Pattern Recognition, 2022, 131: 108889. doi: 10.1016/j.patcog.2022.108889.
[225]	FREDRIKSON M, JHA S, and RISTENPART T. Model inversion attacks that exploit confidence information and basic countermeasures[C]. Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security, Denver, USA, 2015: 1322–1333. doi: 10.1145/2810103.2813677.
[226]	CHEN Huajie, LIU Chi, ZHU Tianqing, et al. When deep learning meets watermarking: A survey of application, attacks and defenses[J]. Computer Standards & Interfaces, 2024, 89: 103830. doi: 10.1016/j.csi.2023.103830.
[227]	DENG Lei, LI Guoqi, HAN Song, et al. Model compression and hardware acceleration for neural networks: A comprehensive survey[J]. Proceedings of the IEEE, 2020, 108(4): 485–532. doi: 10.1109/JPROC.2020.2976475.
[228]	GUO Sensen, LI Xiaoyu, ZHU Peican, et al. ADS-detector: An attention-based dual stream adversarial example detection method[J]. Knowledge-Based Systems, 2023, 265: 110388. doi: 10.1016/j.knosys.2023.110388.
[229]	SCHIOPPA A, FILIPPOVA K, TITOV I, et al. Theoretical and practical perspectives on what influence functions do[C]. Proceedings of the Thirty-Seventh Annual Conference on Neural Information Processing Systems, New Orleans, USA, 2023. (查阅网上资料, 未找到本条文献信息, 请确认) .
[230]	EU Commission. EU Commission—Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain Union legislative acts, COM/2021/206 final[S]. 2021. (查阅网上资料, 未找到本条文献信息, 请确认) .
[231]	ZENG Leya, WANG Buhong, TIAN Jiwei, et al. Threat impact analysis to air traffic control systems through flight delay modeling[J]. Computers & Industrial Engineering, 2021, 162: 107731. doi: 10.1016/j.cie.2021.107731.
[232]	ZENG Leya, WANG Buhong, WANG Tianrui, et al. Research on delay propagation mechanism of air traffic control system based on causal inference[J]. Transportation Research Part C: Emerging Technologies, 2022, 138: 103622. doi: 10.1016/j.trc.2022.103622.
[233]	白洁, 王布宏, 田继伟, 等. 基于动态离散模型的机场ADS-B攻击威胁影响研究[J]. 电光与控制, 2024, 31(7): 27–35,52. doi: 10.3969/j.issn.1671-637X.2024.07.005. BAI Jie, WANG Buhong, TIAN Jiwei, et al. On threat impact of airport ADS-B attack based on dynamic discrete model[J]. Electronics Optics & Control, 2024, 31(7): 27–35,52. doi: 10.3969/j.issn.1671-637X.2024.07.005.

施引文献

期刊类型引用(0)
其他类型引用(1)

资源附件(0)

访问统计

图(15) / 表(13)

计量

文章访问数: 346
HTML全文浏览量: 105
PDF下载量: 70
被引次数: 1

1. 引言
2. 本文模型
2.1 URMNet网络概述
2.2 rvt子模块
2.3 MA模块
2.4 纬度加权损失函数
3. 实验
3.1 数据集与预处理
3.2 实验设置
3.3 MA模块中 $\beta$ 取值对模型性能的影响
3.4 实验对比
3.5 消融实验
3.6 泛化性能对比实验
3.7 复杂度对比实验
4. 结束语

1. 引言
2. 本文模型
2.1 URMNet网络概述
2.2 rvt子模块
2.3 MA模块
2.4 纬度加权损失函数
3. 实验
3.1 数据集与预处理
3.2 实验设置
3.3 MA模块中 $\beta$ 取值对模型性能的影响
3.4 实验对比
3.5 消融实验
3.6 泛化性能对比实验
3.7 复杂度对比实验
4. 结束语

参考文献(233)

施引文献

资源附件(0)

访问统计

姓名
邮箱
手机号码
标题
留言内容
验证码

留言板

空中交通管理系统网络安全研究综述与展望

doi: 10.11999/JEIT240966

通讯作者: 王布宏 wbhgroup@aliyun.com

计量

出版历程

A Review and Prospect of Cybersecurity Research on Air Traffic Management Systems

1. 引言

2. 本文模型

2.1 URMNet网络概述

2.2 rvt子模块

2.2.1 特征图预处理

2.2.2 变换器(Transformer)

2.3 MA模块

2.3.1 SAM

2.3.2 CAM

2.4 纬度加权损失函数

3. 实验

3.1 数据集与预处理

3.2 实验设置

3.3 MA模块中β \beta 取值对模型性能的影响

3.4 实验对比

3.5 消融实验

3.5.1 URMNet网络和损失函数Loss的消融实验

3.5.2 RVT模块和MA模块的消融实验

3.6 泛化性能对比实验

3.7 复杂度对比实验

4. 结束语

期刊类型引用(0)

其他类型引用(1)

计量

出版历程

目录

1. 引言

2. 本文模型

2.1 URMNet网络概述

2.2 rvt子模块

2.3 MA模块

2.4 纬度加权损失函数

3. 实验

3.1 数据集与预处理

3.2 实验设置

3.3 MA模块中β \beta 取值对模型性能的影响

3.4 实验对比

3.5 消融实验

3.6 泛化性能对比实验

3.7 复杂度对比实验

4. 结束语

通讯作者:
王布宏　wbhgroup@aliyun.com

3.3 MA模块中 $\beta$ 取值对模型性能的影响

3.3 MA模块中 $\beta$ 取值对模型性能的影响