Privacy-Preserving Large-Scale AI Models for Intelligent Railway Transportation Systems:Hierarchical Poisoning Attacks and Defenses in Federated Learning

The development of Intelligent Railway Transportation Systems necessitates incorporating privacy-preserving mechanisms into AI models to protect sensitive information and enhance system efficiency.Federated learning offers a promising solution by allowing multiple clients to train models collaboratively without sharing private data.However,despite its privacy benefits,federated learning systems are vulnerable to poisoning attacks,where adversaries alter local model parameters on compromised clients and send malicious updates to the server,potentially compromising the global model’s accuracy.In this study,we introduce PMM(Perturbation coefficient Multiplied by Maximum value),a new poisoning attack method that perturbs model updates layer by layer,demonstrating the threat of poisoning attacks faced by federated learning.Extensive experiments across three distinct datasets have demonstrated PMM’s ability to significantly reduce the global model’s accuracy.Additionally,we propose an effective defense method,namely CLBL(Cluster Layer By Layer).Experiment results on three datasets have confirmed CLBL’s effectiveness.

Artificial-intelligent-powered safety and efficiency improvement for controlling and scheduling in integrated railway systems

The multi-mode integrated railway system,anchored by the high-speed railway,caters to the diverse travel requirements both within and between cities,offering safe,comfortable,punctual,and eco-friendly transportation services.With the expansion of the railway networks,enhancing the efficiency and safety of the comprehensive system has become a crucial issue in the advanced development of railway transportation.In light of the prevailing application of artificial intelligence technologies within railway systems,this study leverages large model technology characterized by robust learning capabilities,efficient associative abilities,and linkage analysis to propose an Artificial-intelligent(AI)-powered railway control and dispatching system.This system is elaborately designed with four core functions,including global optimum unattended dispatching,synergetic transportation in multiple modes,high-speed automatic control,and precise maintenance decision and execution.The deployment pathway and essential tasks of the system are further delineated,alongside the challenges and obstacles encountered.The AI-powered system promises a significant enhancement in the operational efficiency and safety of the composite railway system,ensuring a more effective alignment between transportation services and passenger demands.

A review of artificial intelligence applications in high-speed railway systems

In recent years,the global surge of High-speed Railway(HSR)revolutionized ground transportation,providing secure,comfortable,and punctual services.The next-gen HSR,fueled by emerging services like video surveillance,emergency communication,and real-time scheduling,demands advanced capabilities in real-time perception,automated driving,and digitized services,which accelerate the integration and application of Artificial Intelligence(AI)in the HSR system.This paper first provides a brief overview of AI,covering its origin,evolution,and breakthrough applications.A comprehensive review is then given regarding the most advanced AI technologies and applications in three macro application domains of the HSR system:mechanical manufacturing and electrical control,communication and signal control,and transportation management.The literature is categorized and compared across nine application directions labeled as intelligent manufacturing of trains and key components,forecast of railroad maintenance,optimization of energy consumption in railroads and trains,communication security,communication dependability,channel modeling and estimation,passenger scheduling,traffic flow forecasting,high-speed railway smart platform.Finally,challenges associated with the application of AI are discussed,offering insights for future research directions.

Distributed IRS-Aided DF Relaying Systems:Performance Analysis and Optimization

Intelligent reflecting surface(IRS)is a newly emerged and promising paradigm to substantially improve the performance of wireless communications by constructing favorable communication channels via properly tuning massive reflecting elements.This paper considers a distributed IRS aided decode-and-forward(DF)relaying system over Nakagami-m fading channels.Based on a tight approximation for the distribution of the received signalto-noise ratio(SNR),we first derive exact closed-form expressions of the outage probability,ergodic capacity,and energy efficiency for the considered system.Moreover,we propose the optimal IRS configuration considering the energy efficiency and pilot overhead.Finally,we compare the performance between the distributed IRS-aided DF relaying and multi-IRS-only systems,and verify the analytical results by using monte carlo simulations.

Research on digital twin technology and its application in intelligent operation and maintenance of highspeed railway infrastructure

Purpose–This paper analyzes the application of digital twin technology in the field of intelligent operation and maintenance of high-speed railway infrastructure from the perspective of top-level design.Design/methodology/approach–This paper provides a comprehensive overview of the definition,connotations,characteristics and key technologies of digital twin technology.It also conducts a thorough analysis of the current state of digital twin applications,with a particular focus on the overall requirements for intelligent operation and maintenance of high-speed railway infrastructure.Using the Jinan Yellow River Bridge on the Beijing–Shanghai high-speed railway as a case study,the paper details the construction process of the twin system from the perspectives of system architecture,theoretical definition,model construction and platform design.Findings–Digital twin technology can play an important role in the whole life cycle management,fault prediction and condition monitoring in the field of high-speed rail operation and maintenance.Digital twin technology is of great significance to improve the intelligent level of high-speed railway operation and management.Originality/value–This paper systematically summarizes the main components of digital twin railway.The general framework of the digital twin bridge is given,and its application in the field of intelligent operation and maintenance is prospected.

System architecture and basic platform for intelligent high-speed railway

Purpose-This paper aims to provide top-level design and basic platform for intelligent application in China high-speed railway.Design/methodology/approach-Based on the analysis for the future development trends of world railway,combined with the actual development needs in China high-speed railway,The definition and scientific connotation of intelligent high-speed railway(IHSR)are given at first,and then the system architecture of IHSR are outlined,including 1 basic platform,3 business sectors,10 business fields,and 18 innovative applications.At last,a basic platform with cloud edge integration for IHSR is designed.Findings-The rationality,feasibility and implementability of the system architecture of IHSR have been verified on and applied to the Beijing-Zhangjiakou high-speed railway,providing important support for the construction and operation of the world’s first IHSR.Originality/value-This paper systematically gives the definition and connotation of the IHSR and put forward the system architecture of IHSR for first time.It will play the most important role in the design,construction and operation of IHSR.

Research on Active Safety Methodologies for Intelligent Railway Systems

Safety is essential when building a strong transportation system.As a key development direction in the global railway system,the intelligent railway has safety at its core,making safety a top priority while pursuing the goals of efficiency,convenience,economy,and environmental friendliness.This paper describes the state of the art and proposes a system architecture for intelligent railway systems.It also focuses on the development of railway safety technology at home and abroad,and proposes the active safety method and technology system based on advanced theoretical methods such as the in-depth integration of cyber–physical systems(CPS),data-driven models,and intelligent computing.Finally,several typical applications are demonstrated to verify the advancement and feasibility of active safety technology in intelligent railway systems.

The Study on the Framework and Implementation Scheme of the Railway Intelligent Cold Chain Logistics System

With the development of intelligent and communication technology, traditional logistics has gradually transformed into intelligent logistics. The construction of the railway intelligent cold chain logistics system is an effective measure to implement the structural adjustment of supply, which conforms to the times. Firstly, basic features and design principles of the railway intelligent cold chain logistics system were described. Secondly, target functions and the system framework of the railway intelligent cold chain logistics were put forward, which takes information service as a foundation, management service as a guarantee, business capability as kernel and expansion capability as support. Subsequently, the technical approaches were studied at the hardware, processing and software levels. Finally, development strategies of the railway intelligent cold chain logistics were discussed, including optimizing the layout of infrastructure, improving the performance of equipment, extending the reach of services, deepening the cooperation of technologies and promoting the formulation of standards which provides feasible references for the railway cold chain logistics to explore modern business models.

IRS辅助的车联网相移设计和信道对齐策略

针对车联网中车辆的动态移动以及信号的随机散射所导致的信道快速变化问题,研究了智能反射面(intelligent reflecting surface,IRS)辅助的车联网通信并提出了联合相移设计和信道对齐策略,以削减车联网中多普勒频移的影响并提升通信性能。提出的策略由两个阶段构成。在第一阶段设计IRS的相移以改善级联信道的衰落状态,在策略的第二阶段进行修正函数的设计,实现了直接信道和级联信道的信道对齐,从而减少直接信道快衰落状态对通信性能的影响。通过联合相移设计和信道对齐策略中设计的反射相移以及修正函数,实现了通信性能的提升。仿真结果表明,所提策略相比于相移优化策略可以提升8.8%的频谱效率。

城市街道下场景IRS辅助D2D通信系统波束成形设计

针对城市街道场景下蜂窝用户和D2D通信用户共享频谱以及城市街道下无线信道特性,该文提出一种IRS辅助的联合波束成形设计方法。在D2D链路信号与干扰加噪声比的约束下,以最大化蜂窝用户容量为目标,设计了最优的波束形成向量、相移矩阵和D2D链路发射功率。引入松弛变量将非凸且变量耦合的优化问题转换为解耦后的凸优化问题和二分法搜索功率分配,采用黎曼共轭梯度算法对反射相移矩阵进行优化。仿真结果表明,所提算法收敛性较好,且与基准方案相比能有效地提升用户信道容量。

利用元学习算法的IRS-OTFS通信系统信道估计

针对高多普勒场景下智能反射表面(IRS)辅助多用户通信系统存在的信道估计传输开销大的问题,该文结合正交时频空间(OTFS)调制特点构造一种IRS-OTFS通信系统,充分发挥IRS和OTFS的性能优势,并在此基础上提出一种学习率自适应的模型无关元学习(MAML)算法。对IRS-OTFS多用户信道估计任务做离线训练,根据各任务的收敛速度自适应地调整学习率,防止训练失衡,并利用信道之间的相关性和MAML算法的少样本、泛化特性得到全局模型和适应性模型,快速学习新用户信道的传输特性,降低传输开销,提高信道估计准确性。理论分析和仿真结果表明,该算法在信道传输条件相同的情况下,将传输开销降低了大约50%,并相对于基准算法有4.8 dB左右的性能提升。

IRS辅助去蜂窝mMIMO系统低复杂度波束赋形算法

针对目前智能反射面(intelligent reflecting surface,IRS)辅助去蜂窝大规模多输入多输出(massive multiple-input multiple-output,mMIMO)系统现有算法复杂度高的问题,提出了一种基于IRS辅助去蜂窝mMIMO系统联合波束赋形低复杂度算法。在满足还原用户接收的信号、IRS相移和发射功率约束的条件下,以得到最优发射波束赋形向量为目标,构建了一个联合优化接入点(access point,AP)主动波束赋形和IRS被动波束赋形问题。利用交替优化方法将原问题转换为两个易于处理的子问题,并基于解非齐次线性方程组,采用Lagrange乘数法和人工蜂群算法对子问题进行求解。仿真结果表明,相较于传统算法,所提算法有更低的复杂度和误码率,可使用更小的发射功率和频谱效率等资源获得与其他方案相当的性能。

IRS-Aided SWIPT Systems with Power Splitting and Artificial Noise

Intelligent reflecting surface(IRS) is regarded as a promising technology because it can achieve higher passive beamforming gain. In particular, the IRS assisted simultaneous wireless information and power transfer(SWIPT) system can make the information decoding receivers(IDRs) have a higher signal-to-noise ratio(SNR), and the energy harvesting receivers(EHRs) have the guarantee of minimum harvested energy threshold. Motivated by the above,in this paper, we use the power splitting(PS) at the user and introduce artificial noise(AN) into the access point(AP), so that the user in system can harvest energy and decode information simultaneously,further improve the security of user. We jointly optimize the beamforming matrix at AP, the reflection phase shift at IRS and the PS ratio, in order to maximize the user’s achievable secrecy rate, subject to the user’s minimum harvested energy threshold and AP’s transmission power. Due to the introduction of PS ratio, the coupling between variables is increased,and the complexity of the problem is significantly increased. Furthermore, the problem is non-convex, so we propose an efficient algorithm based on Taylor Formula, semi-definite relaxation(SDR) and alternating optimization(AO) to get the solution. Numerical results show that the proposed IRS-SWIPT system with PS and AN achieves significant performance improvement compared with other benchmark scheme.

Current state and predicted technological trends in global railway intelligent digital transformation

Purpose–In recent years,railway systems worldwide have faced challenges such as the modernization of engineering projects,efficient management of intelligent digital railway equipment,rapid growth in passenger and freight transport demands,customized transport services and ubiquitous transport safety.The transformation toward intelligent digital transformation in railways has emerged as an effective response to the formidable challenges confronting the railway industry,thereby becoming an inevitable global trend in railway development.Design/methodology/approach–This paper,therefore,conducts a comprehensive analysis of the current state of global railway intelligent digital transformation,focusing on the characteristics and applications of intelligent digital transformation technology.It summarizes and analyzes relevant technologies and applicable scenarios in the realm of railway intelligent digital transformation,theoretically elucidating the development process of global railway intelligent digital transformation and,in practice,providing guidance and empirical examples for railway intelligence and digital transformation.Findings–Digital and intelligent technologies follow a wave-like pattern of continuous iterative evolution,progressing from the early stages,to a period of increasing attention and popularity,then to a phase of declining interest,followed by a resurgence and ultimately reaching a mature stage.Originality/value–The results offer reference and guidance to fully leverage the opportunities presented by the latest wave of the digitalization revolution,accelerate the overall upgrade of the railway industry and promote global collaborative development in railway intelligent digital transformation.

Key technologies and applications of intelligent dispatching command for high-speed railway in China

Purpose–The intelligent Central Traffic Control(CTC)system plays a vital role in establishing an intelligent high-speed railway(HSR)system.As the core of HSR transportation command,the intelligent CTC system is a new HSR dispatching command system that integrates the widely used CTC in China with the practical service requirements of intelligent dispatching.This paper aims to propose key technologies and applications for intelligent dispatching command in HSR in China.Design/methodology/approach–This paper first briefly introduces the functions and configuration of the intelligent CTC system.Some new servers,terminals and interfaces are introduced,which are plan adjustment server/terminal,interface for automatic train operation(ATO),interface for Dynamic Monitoring System of Train Control Equipment(DMS),interface for Power Supervisory Control and Data Acquisition(PSCADA),interface for Disaster Monitoring,etc.Findings–The key technologies applied in the intelligent CTC system include automatic adjustment of train operation plans,safety control of train routes and commands,traffic information data platform,integrated simulation of traffic dispatching and ATO function.These technologies have been applied in the Beijing-Zhangjiakou HSR,which commenced operations at the end of 2019.Implementing these key intelligent functions has improved the train dispatching command capacity,ensured the safe operation of intelligent HSR,reduced the labor intensity of dispatching operators and enhanced the intelligence level of China’s dispatching system.Originality/value–This paper provides further challenges and research directions for the intelligent dispatching command of HSR.To achieve the objectives,new measures need to be conducted,including the development of advanced technologies for intelligent dispatching command,coping with new requirements with the development of China’s railway signaling system,the integration of traffic dispatching and train control and the application of AI and data-driven modeling and methods.

基于不动点深度学习的IRS辅助毫米波移动通信系统全信道估计

将智能反射面(IRS)与大规模MIMO结合能够保证和提高毫米波通信系统性能。针对基站(BS)-用户直连信道与用户-IRS-BS反射信道混叠场景,该文提出一种自适应的全信道估计方法。首先,引入辅助变量,采用原子范数将直连信道与反射信道的稀疏角度域子空间进行关联;然后,利用原子范数最小化将全信道估计问题建模为连续角度域稀疏矩阵重建规划;最后,基于不动点深度学习网络设计低复杂度的问题求解算法。该算法不仅能够克服传统基于模型解法中非线性估计算子对先验知识的依赖还可根据移动场景变化自适应调节算法复杂度。仿真结果表明,所提算法能够避免传统时分估计策略引起的差错传播效应,具有更高的估计精度和更低的复杂度。

Channel Capacity Optimization Based on Riemannian Trust Region Algorithm in IRS-Aided Communication System

Intelligent Reflecting Surface (IRS) can offer unprecedented channel capacity gains since it can reconfigure the signal propagation environment. We decide to maximize the channel capacity by jointly optimizing the transmit-power-constrained precoding matrix at the base station and the unit-modulus-constrained phase shift vector at the IRS in IRS-assisted multi-user downlink communication. We first convert the resulting non-convex problem into an equivalent problem, then use the alternate optimization algorithm. While fixing the phase shift vector, we can obtain the optimal precoding matrix directly by adopting standard optimization packages. While fixing the precoding matrix, we propose the Riemannian Trust-Region (RTR) algorithm to solve this optimization problem. And the key of the RTR algorithm is the solution of the trust-region sub-problem. We first adopt the accurate solution based on Newton's (ASNT) method to solve this sub-problem, which can obtain the global solution but cannot guarantee that the solution is optimal since the initial iteration point is difficult to choose. Then, we propose the Improved-Polyline (IPL) method, which can avoid the difficulty of the ASNT method and improve convergence speed and calculation efficiency. The numerical results show that the RTR algorithm has more significant performance gains and faster convergence speed compared with the existing approaches.

IRS Assisted UAV Communications against Proactive Eavesdropping in Mobile Edge Computing Networks

In this paper,we consider mobile edge computing(MEC)networks against proactive eavesdropping.To maximize the transmission rate,IRS assisted UAV communications are applied.We take the joint design of the trajectory of UAV,the transmitting beamforming of users,and the phase shift matrix of IRS.The original problem is strong non-convex and difficult to solve.We first propose two basic modes of the proactive eavesdropper,and obtain the closed-form solution for the boundary conditions of the two modes.Then we transform the original problem into an equivalent one and propose an alternating optimization(AO)based method to obtain a local optimal solution.The convergence of the algorithm is illustrated by numerical results.Further,we propose a zero forcing(ZF)based method as sub-optimal solution,and the simulation section shows that the proposed two schemes could obtain better performance compared with traditional schemes.

Intelligent reflecting surface-assisted cognitive radio-inspired rate-splitting multiple access systems

Intelligent reflecting surface(IRS)has been widely regarded as a promising technology for configuring wireless propagation environments.In this paper,we utilize IRS to assist transmission of a secondary user(SU)in a cognitive radio-inspired rate-splitting multiple access(CR-RSMA)system in which a primary user's(PU's)quality of service(QoS)requirements must be guaranteed.Without introducing intolerable interference to deteriorate the PU's outage performance,the SU conducts rate-splitting to transmit its signal to the base-station through the direct link and IRS reflecting channels.For the IRS-assisted CR-RSMA(IRS-CR-RSMA)scheme,we derive the optimal transmit power allocation,target rate allocation,and successive interference cancellation decoding order to enhance the outage performance of the SU.The closed-form expression for the SU's outage probability achieved by the IRS-CR-RSMA scheme is derived.Various simulation results are presented to clarify the enhanced outage performance achieved by the proposed IRS-CR-RSMA scheme over the CR-RSMA scheme.

基于RSMA的IRS辅助认知无线网络安全波束成形方案

研究对象为一种结合了认知无线电网络、智能反射面(Intelligent Reflecting Surface,IRS)以及速率分割多址(Rate Splitting Multiple Access,RSMA)技术的IRS辅助的认知无线网络,通过充分利用IRS和RSMA的优势,该网络展现了在提高频谱效率方面的潜在巨大优势。然而,该网络容易受到安全问题的困扰,多用户共享相同频谱会增加干扰和隐私泄露的风险。为实现频谱效率和安全效率的联合考量,提出了基于RSMA的主次网络联合安全波束成形设计方法,以充分利用次用户产生的干扰,来抑制潜在窃听者,并在安全和能效之间取得平衡。在此框架下,研究了考虑多种约束条件下的主用户安全能效优化问题,并通过交替迭代波束成形的设计方案,结合连续凸近似和泰勒展开方法,实现了有效求解。该方法在基于RSMA的IRS辅助的认知无线网络中能充分利用次网络发射机的干扰,以提升主网络的安全性能,最终通过联合安全波束成形设计,实现了传输效率和安全性能的有效权衡。