Journal of Information Security and Applications最新文献_第4页

Knit scrambling: A novel image scrambling framework and its demonstration in image encryption 编织置乱：一种新的图像置乱框架及其在图像加密中的应用

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-15 DOI: 10.1016/j.jisa.2025.104326

Muhammed Saadetdin KAYA , Kenan İNCE

The exponential growth of visual data and the expansion of resource-constrained IoT platforms have intensified the demand for lightweight yet secure image protection schemes. Conventional ciphers, while cryptographically strong, often fail to meet real-time and hardware-efficiency requirements for image data. To address this gap, this study presents the Knit Scrambling (KS) framework, a textile-inspired deterministic permutation framework designed for reversible image scrambling with linear computational cost. This approach models an image as a sequence interwoven from multiple subsequences following cyclic knitting patterns, ensuring both reversibility and high diffusion. A specific instantiation, termed Triple Check Pattern (TCP), realizes the KS framework by dividing the image into three subsequences and applying cyclic pattern rotations to enhance pixel decorrelation while preserving strict invertibility. The confusion process is integrated with a lightweight diffusion stage based on a key-nonce-derived chaotic keystream generated by a one-dimensional logistic map, eliminating plaintext dependence and enabling per-image uniqueness. Experimental analyses conducted on benchmark color images show near-uniform histograms, high entropy close to eight bits, and strong differential performance, with average NPCR around 99.6 percent and UACI approximately 33.5 percent. Statistical randomness evaluation using the NIST SP 800-22 test suite confirms the scheme’s ability to produce unpredictable ciphertexts, while runtime benchmarking on both desktop and embedded-class hardware demonstrates real-time feasibility. The results indicate that the proposed framework provides an effective and hardware-efficient alternative to existing chaos-based and geometric scrambling approaches for lightweight image encryption in IoT environments. The proposed framework (KS) defines a general textile-inspired permutation model, while its implementation through the TCP algorithm demonstrates how this model can be practically realized to achieve efficient and reversible image scrambling.

视觉数据的指数级增长和资源受限的物联网平台的扩展加剧了对轻量级但安全的图像保护方案的需求。传统的密码虽然密码学很强，但往往不能满足图像数据的实时性和硬件效率要求。为了解决这一差距，本研究提出了针织置乱（KS）框架，这是一种受纺织品启发的确定性排列框架，设计用于具有线性计算成本的可逆图像置乱。这种方法将图像建模为由多个子序列按照循环编织模式交织而成的序列，确保了可逆性和高扩散性。一个具体的实例，称为三重检查模式（TCP），通过将图像划分为三个子序列并应用循环模式旋转来实现KS框架，以增强像素去相关，同时保持严格的可逆性。混淆过程与基于一维逻辑映射生成的键非派生混沌密钥流的轻量级扩散阶段相集成，消除了明文依赖并实现了每个图像的唯一性。对基准彩色图像进行的实验分析显示，直方图接近均匀，高熵接近8位，差异性能强，平均NPCR约为99.6%，UACI约为33.5%。使用NIST SP 800-22测试套件的统计随机性评估证实了该方案产生不可预测的密文的能力，而在桌面和嵌入式类硬件上的运行时基准测试证明了实时可行性。结果表明，所提出的框架为物联网环境中的轻量级图像加密提供了一种有效且硬件效率高的替代方案，可以替代现有的基于混沌和几何置乱的方法。提出的框架（KS）定义了一个通用的纺织品启发的排列模型，而通过TCP算法的实现演示了该模型如何实际实现，以实现高效和可逆的图像置乱。

{"title":"Knit scrambling: A novel image scrambling framework and its demonstration in image encryption","authors":"Muhammed Saadetdin KAYA , Kenan İNCE","doi":"10.1016/j.jisa.2025.104326","DOIUrl":"10.1016/j.jisa.2025.104326","url":null,"abstract":"<div><div>The exponential growth of visual data and the expansion of resource-constrained IoT platforms have intensified the demand for lightweight yet secure image protection schemes. Conventional ciphers, while cryptographically strong, often fail to meet real-time and hardware-efficiency requirements for image data. To address this gap, this study presents the Knit Scrambling (KS) framework, a textile-inspired deterministic permutation framework designed for reversible image scrambling with linear computational cost. This approach models an image as a sequence interwoven from multiple subsequences following cyclic knitting patterns, ensuring both reversibility and high diffusion. A specific instantiation, termed Triple Check Pattern (TCP), realizes the KS framework by dividing the image into three subsequences and applying cyclic pattern rotations to enhance pixel decorrelation while preserving strict invertibility. The confusion process is integrated with a lightweight diffusion stage based on a key-nonce-derived chaotic keystream generated by a one-dimensional logistic map, eliminating plaintext dependence and enabling per-image uniqueness. Experimental analyses conducted on benchmark color images show near-uniform histograms, high entropy close to eight bits, and strong differential performance, with average NPCR around 99.6 percent and UACI approximately 33.5 percent. Statistical randomness evaluation using the NIST SP 800-22 test suite confirms the scheme’s ability to produce unpredictable ciphertexts, while runtime benchmarking on both desktop and embedded-class hardware demonstrates real-time feasibility. The results indicate that the proposed framework provides an effective and hardware-efficient alternative to existing chaos-based and geometric scrambling approaches for lightweight image encryption in IoT environments. The proposed framework (KS) defines a general textile-inspired permutation model, while its implementation through the TCP algorithm demonstrates how this model can be practically realized to achieve efficient and reversible image scrambling.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104326"},"PeriodicalIF":3.7,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145791475","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Lightweight orthogonal perturbation for privacy-preserving federated learning against poisoning attacks 针对中毒攻击的保护隐私的联邦学习轻量级正交摄动

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-15 DOI: 10.1016/j.jisa.2025.104345

Chuanyu Peng , Hequn Xian

Federated Learning (FL) is a distributed machine learning approach where multiple users collaboratively train a shared model without sharing raw data, thereby preserving data privacy. However, FL remains vulnerable to inference and poisoning attacks, which can compromise privacy and degrade global model performance. Therefore, many privacy-preserving frameworks have been proposed. Among these, mask-based frameworks provide advantages in efficiency and flexibility, but are particularly susceptible to poisoning attacks by malicious users. To overcome this challenge, we propose LOPAS-FL, an efficient, privacy-preserving, and robust mask-based federated learning scheme. It first introduces a gradient-splitting and orthogonal perturbation mechanism to ensure privacy through computational indistinguishability. Meanwhile, a dual-server architecture conducts multi-dimensional verification across gradient direction, distribution, and homogeneity. Only gradients that pass all validations are aggregated. This approach effectively defends against poisoning attacks and ensures the quality and robustness of the final model. Security analysis and experiments show that LOPAS-FL effectively detects and mitigates poisoning attacks, outperforming existing approaches in efficiency.

联邦学习（FL）是一种分布式机器学习方法，其中多个用户协作训练共享模型，而不共享原始数据，从而保护数据隐私。然而，FL仍然容易受到推理和中毒攻击，这可能会损害隐私并降低全局模型的性能。因此，人们提出了许多隐私保护框架。其中，基于掩码的框架在效率和灵活性方面具有优势，但特别容易受到恶意用户的中毒攻击。为了克服这一挑战，我们提出了一种高效、隐私保护和鲁棒的基于掩码的联邦学习方案LOPAS-FL。它首先引入了梯度分裂和正交摄动机制，通过计算不可区分来确保隐私。同时，双服务器架构跨梯度方向、分布和同质性进行多维度验证。只有通过所有验证的梯度才被聚合。这种方法有效地防御了中毒攻击，并确保了最终模型的质量和健壮性。安全分析和实验表明，LOPAS-FL可以有效检测和减轻中毒攻击，效率优于现有方法。

{"title":"Lightweight orthogonal perturbation for privacy-preserving federated learning against poisoning attacks","authors":"Chuanyu Peng , Hequn Xian","doi":"10.1016/j.jisa.2025.104345","DOIUrl":"10.1016/j.jisa.2025.104345","url":null,"abstract":"<div><div>Federated Learning (FL) is a distributed machine learning approach where multiple users collaboratively train a shared model without sharing raw data, thereby preserving data privacy. However, FL remains vulnerable to inference and poisoning attacks, which can compromise privacy and degrade global model performance. Therefore, many privacy-preserving frameworks have been proposed. Among these, mask-based frameworks provide advantages in efficiency and flexibility, but are particularly susceptible to poisoning attacks by malicious users. To overcome this challenge, we propose LOPAS-FL, an efficient, privacy-preserving, and robust mask-based federated learning scheme. It first introduces a gradient-splitting and orthogonal perturbation mechanism to ensure privacy through computational indistinguishability. Meanwhile, a dual-server architecture conducts multi-dimensional verification across gradient direction, distribution, and homogeneity. Only gradients that pass all validations are aggregated. This approach effectively defends against poisoning attacks and ensures the quality and robustness of the final model. Security analysis and experiments show that LOPAS-FL effectively detects and mitigates poisoning attacks, outperforming existing approaches in efficiency.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104345"},"PeriodicalIF":3.7,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145791476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An imperceptible dynamic anticipated backdoor attack in federated learning 联邦学习中不可察觉的动态预期后门攻击

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-11 DOI: 10.1016/j.jisa.2025.104317

Yingqiang Xie , Wei Ren , Tianqing Zhu , Lianchong Zhang

Federated Learning is a distributed machine learning paradigm that allows multiple clients to collaboratively train a global model while preserving privacy by avoiding the exchange of raw data. However, its distributed nature makes it vulnerable to backdoor attacks, which threaten the integrity and security of the model. Existing attacks often rely on fixed triggers or optimizations of the local model, failing to adapt to dynamic updates of the global model. We propose a new and effective attack named IDABA (Imperceptible Dynamic Anticipated Backdoor Attack), a novel dynamic backdoor attack method for FL, addressing these limitations by ensuring visual imperceptibility and persistence. IDABA generates visually imperceptible poisoned samples and employs Model-Contrastive Loss (MOON) to maintain similarity with the global model. It also predicts future global model states to optimize trigger effectiveness. Experiments on CIFAR10, MNIST, GTSRB, and TinyImageNet show that IDABA achieves higher Attack Success Rates (ASR) while maintaining model accuracy. It demonstrates strong adaptability against defense mechanisms such as Krum and Multi-Krum. GradCam analysis and image quality metrics confirm the visual stealthiness of IDABA’s backdoor samples.

联邦学习是一种分布式机器学习范例，它允许多个客户端协作训练全局模型，同时通过避免交换原始数据来保护隐私。然而，它的分布式特性使其容易受到后门攻击，从而威胁到模型的完整性和安全性。现有的攻击通常依赖于固定的触发器或局部模型的优化，无法适应全局模型的动态更新。我们提出了一种新的有效的攻击方法IDABA （imvisible Dynamic预期后门攻击），它通过保证视觉的不可感知性和持久性来解决这些限制。IDABA产生视觉上难以察觉的中毒样本，并使用模型对比损失（MOON）来保持与全局模型的相似性。它还预测未来的全局模型状态，以优化触发器的有效性。在CIFAR10、MNIST、GTSRB和TinyImageNet上的实验表明，IDABA在保持模型精度的同时实现了更高的攻击成功率（ASR）。它对克鲁姆和多克鲁姆等防御机制具有较强的适应性。GradCam分析和图像质量指标证实了IDABA后门样本的视觉隐身性。

{"title":"An imperceptible dynamic anticipated backdoor attack in federated learning","authors":"Yingqiang Xie , Wei Ren , Tianqing Zhu , Lianchong Zhang","doi":"10.1016/j.jisa.2025.104317","DOIUrl":"10.1016/j.jisa.2025.104317","url":null,"abstract":"<div><div>Federated Learning is a distributed machine learning paradigm that allows multiple clients to collaboratively train a global model while preserving privacy by avoiding the exchange of raw data. However, its distributed nature makes it vulnerable to backdoor attacks, which threaten the integrity and security of the model. Existing attacks often rely on fixed triggers or optimizations of the local model, failing to adapt to dynamic updates of the global model. We propose a new and effective attack named IDABA (Imperceptible Dynamic Anticipated Backdoor Attack), a novel dynamic backdoor attack method for FL, addressing these limitations by ensuring visual imperceptibility and persistence. IDABA generates visually imperceptible poisoned samples and employs Model-Contrastive Loss (MOON) to maintain similarity with the global model. It also predicts future global model states to optimize trigger effectiveness. Experiments on CIFAR10, MNIST, GTSRB, and TinyImageNet show that IDABA achieves higher Attack Success Rates (ASR) while maintaining model accuracy. It demonstrates strong adaptability against defense mechanisms such as Krum and Multi-Krum. GradCam analysis and image quality metrics confirm the visual stealthiness of IDABA’s backdoor samples.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104317"},"PeriodicalIF":3.7,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738766","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enabling diverse styles coverless image steganography with two-stage latent transformation and diffusion model 利用两级潜伏变换和扩散模型实现多种风格的无盖图像隐写

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-11 DOI: 10.1016/j.jisa.2025.104328

Jiangtao Guo, Buwei Tian, Junyong Jiang, Lu Dong

Coverless image steganography (CIS) aims to map secret images into container images without modifying the original images for concealment purposes. However, most current CIS methods rely heavily on deep learning models, which require extensive training datasets and demonstrate limited robustness to variations in image styles. Particularly when applied to images with substantial stylistic variations, these methods often produce unsatisfactory steganographic results, leading to significant degradation in the quality of steganographic image (stego-image). Furthermore, existing diffusion model-based CIS approaches can only achieve effective concealment between images with similar styles, thereby limiting the diversity of application scenarios. To address these limitations, we propose a training-free CIS method based on the diffusion model (DStyleStego), which does not rely on the traditional training, and can effectively handle different styles of images, guaranteeing the image quality and the security of steganographic information. Specifically, we design a two-stage latent transformation method to improve the security and flexibility of image steganography. In addition, we introduce a detail compensation function to recover detail information lost during the diffusion process to improve the quality and fidelity of the generated images. Extensive experimental results demonstrate that DStyleStego achieves efficient and stable steganography across diverse image datasets (Stego260 and UniStega) while exhibiting significant advantages in terms of image quality preservation.

无覆盖图像隐写术（CIS）的目的是在不修改原始图像的情况下，将秘密图像映射到容器图像中，以达到隐藏的目的。然而，目前大多数CIS方法严重依赖于深度学习模型，这需要大量的训练数据集，并且对图像风格变化的鲁棒性有限。特别是当应用于具有大量风格变化的图像时，这些方法通常会产生令人不满意的隐写结果，导致隐写图像（隐写图像）质量的显着下降。此外，现有的基于扩散模型的CIS方法只能在样式相似的图像之间实现有效的隐藏，限制了应用场景的多样性。针对这些局限性，我们提出了一种基于扩散模型的无训练CIS方法（DStyleStego），该方法不依赖于传统的训练，可以有效地处理不同风格的图像，保证了图像质量和隐写信息的安全性。具体来说，为了提高图像隐写的安全性和灵活性，我们设计了一种两阶段隐写变换方法。此外，我们引入了细节补偿函数来恢复扩散过程中丢失的细节信息，以提高生成图像的质量和保真度。大量的实验结果表明，DStyleStego在不同的图像数据集（Stego260和UniStega）上实现了高效和稳定的隐写，同时在图像质量保持方面表现出显著的优势。

{"title":"Enabling diverse styles coverless image steganography with two-stage latent transformation and diffusion model","authors":"Jiangtao Guo, Buwei Tian, Junyong Jiang, Lu Dong","doi":"10.1016/j.jisa.2025.104328","DOIUrl":"10.1016/j.jisa.2025.104328","url":null,"abstract":"<div><div>Coverless image steganography (CIS) aims to map secret images into container images without modifying the original images for concealment purposes. However, most current CIS methods rely heavily on deep learning models, which require extensive training datasets and demonstrate limited robustness to variations in image styles. Particularly when applied to images with substantial stylistic variations, these methods often produce unsatisfactory steganographic results, leading to significant degradation in the quality of steganographic image (stego-image). Furthermore, existing diffusion model-based CIS approaches can only achieve effective concealment between images with similar styles, thereby limiting the diversity of application scenarios. To address these limitations, we propose a training-free CIS method based on the diffusion model (DStyleStego), which does not rely on the traditional training, and can effectively handle different styles of images, guaranteeing the image quality and the security of steganographic information. Specifically, we design a two-stage latent transformation method to improve the security and flexibility of image steganography. In addition, we introduce a detail compensation function to recover detail information lost during the diffusion process to improve the quality and fidelity of the generated images. Extensive experimental results demonstrate that DStyleStego achieves efficient and stable steganography across diverse image datasets (Stego260 and UniStega) while exhibiting significant advantages in terms of image quality preservation.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104328"},"PeriodicalIF":3.7,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738685","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A class-imbalance-aware intrusion detection system based on spatiotemporal graph neural networks for software-defined vehicles 基于时空图神经网络的软件定义车辆类不平衡感知入侵检测系统

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-11 DOI: 10.1016/j.jisa.2025.104340

Sishan Wang , Youqun Zhao , Xin Fu , Huachao Si , Wentao Wang , Lei Xue

Advanced driving systems (ADS) and vehicle-to-everything (V2X) technologies are accelerating the shift to software-defined vehicles (SDVs), which have dramatically increased the demand for connectivity and bandwidth of in-vehicle networks (IVNs), e.g., automotive Ethernet. The SOME/IP (Scalable service-Oriented MiddlewarE over IP) protocol, a middleware standard for automotive Ethernet, presents unique security challenges: the lack of security mechanisms and its dynamic session behaviors render traditional rule-based Intrusion Detection Systems (IDSs) ineffective. To address these challenges, we propose GATransformer, a novel hybrid Graph Attention Network (GAT) with Transformer architecture that can learn spatial-temporal dependencies for SOME/IP-based IVNs in the class imbalance scenario. A comprehensive evaluation on a SOME/IP dataset built from a real-world Connected and Autonomous Vehicle (CAV) indicates that the GATransformer enhances robustness in a class imbalance scenario and significantly outperforms conventional Deep Learning (DL) architectures, including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), standalone Transformers, and baseline GAT. The proposed model achieves an F1-score of 0.9999 with 0.004 ms inference latency on server-grade hardware (NVIDIA RTX 3090) and maintains robust performance (F1-score: 0.9983) with sub-millisecond latency (0.151 ms) when deployed on an automotive-grade embedded platform (NVIDIA Jetson Orin Nano). These results validate the possibility of deploying the hybrid Graph Neural Networks (GNNs) for real-time automotive intrusion detection, representing a significant advancement toward securing next-generation service-oriented architectures (SOA) against evolving cyber threats.

先进驾驶系统（ADS）和车联网（V2X）技术正在加速向软件定义车辆（sdv）的转变，这极大地增加了对车载网络（ivn）（如汽车以太网）的连接性和带宽的需求。SOME/IP（基于IP的可伸缩面向服务的中间件）协议是汽车以太网的中间件标准，它提出了独特的安全挑战：缺乏安全机制及其动态会话行为使得传统的基于规则的入侵检测系统（ids）无效。为了解决这些挑战，我们提出了一种具有Transformer架构的新型混合图注意网络（GAT） gattransformer，它可以在类不平衡场景中学习基于SOME/ ip的ivn的时空依赖关系。对现实世界联网和自动驾驶汽车（CAV）构建的SOME/IP数据集的综合评估表明，gattransformer增强了类不平衡场景中的鲁棒性，并显著优于传统的深度学习（DL）架构，包括卷积神经网络（CNN）、长短期记忆（LSTM）、独立变压器和基线GAT。该模型在服务器级硬件（NVIDIA RTX 3090）上实现了0.9999的f1分数和0.004 ms的推理延迟，并在部署在汽车级嵌入式平台（NVIDIA Jetson Orin Nano）上时保持了亚毫秒级延迟（0.151 ms）的稳健性能（f1分数：0.9983）。这些结果验证了在实时汽车入侵检测中部署混合图神经网络（gnn）的可能性，代表了在保护下一代面向服务的架构（SOA）免受不断发展的网络威胁方面取得的重大进展。

{"title":"A class-imbalance-aware intrusion detection system based on spatiotemporal graph neural networks for software-defined vehicles","authors":"Sishan Wang , Youqun Zhao , Xin Fu , Huachao Si , Wentao Wang , Lei Xue","doi":"10.1016/j.jisa.2025.104340","DOIUrl":"10.1016/j.jisa.2025.104340","url":null,"abstract":"<div><div>Advanced driving systems (ADS) and vehicle-to-everything (V2X) technologies are accelerating the shift to software-defined vehicles (SDVs), which have dramatically increased the demand for connectivity and bandwidth of in-vehicle networks (IVNs), e.g., automotive Ethernet. The SOME/IP (Scalable service-Oriented MiddlewarE over IP) protocol, a middleware standard for automotive Ethernet, presents unique security challenges: the lack of security mechanisms and its dynamic session behaviors render traditional rule-based Intrusion Detection Systems (IDSs) ineffective. To address these challenges, we propose GATransformer, a novel hybrid Graph Attention Network (GAT) with Transformer architecture that can learn spatial-temporal dependencies for SOME/IP-based IVNs in the class imbalance scenario. A comprehensive evaluation on a SOME/IP dataset built from a real-world Connected and Autonomous Vehicle (CAV) indicates that the GATransformer enhances robustness in a class imbalance scenario and significantly outperforms conventional Deep Learning (DL) architectures, including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), standalone Transformers, and baseline GAT. The proposed model achieves an F1-score of 0.9999 with 0.004 ms inference latency on server-grade hardware (NVIDIA RTX 3090) and maintains robust performance (F1-score: 0.9983) with sub-millisecond latency (0.151 ms) when deployed on an automotive-grade embedded platform (NVIDIA Jetson Orin Nano). These results validate the possibility of deploying the hybrid Graph Neural Networks (GNNs) for real-time automotive intrusion detection, representing a significant advancement toward securing next-generation service-oriented architectures (SOA) against evolving cyber threats.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104340"},"PeriodicalIF":3.7,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Improved LSB substitution based semi-blind fragile watermarking for high-accuracy tamper localization 基于改进LSB替换的半盲脆弱水印高精度篡改定位

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-11 DOI: 10.1016/j.jisa.2025.104330

Pengfei Li , Weijia Cao , Tao Yu

Fragile watermarking is an effective technique for detecting image tampering, yet current block-based fragile watermarking methods often struggle with limited watermarked image quality and high false-positive rates. To overcome these limitations, this paper proposes an enhanced semi-blind fragile watermarking method based on an improved least significant bit (LSB) substitution (ILSBS), named TLFW. First, the original image is divided into non-overlapping blocks, within which a novel rule-based ILSBS approach dynamically adjusts the watermark embedding bits to minimize pixel distortion effectively. This significantly enhances the visual quality of the watermarked image. Next, a tamper localization optimization (TLO) strategy is introduced to substantially reduce false positives by refining detection results around reference points. Extensive experimental results demonstrate that the TLFW approach improves the peak signal-to-noise ratio (PSNR) of watermarked images from 44 dB to 46 dB, consistently reduces the false positive rate (FPR) across multiple attack scenarios while achieving 45 % FPR reduction and 52 % tamper detection rate (TDR) gain specifically for the text-addition attack, and lowers computational costs significantly compared to existing methods. The proposed TLFW scheme is compatible with both grayscale and color images and does not require the original image during watermark extraction, making it highly suitable for practical image authentication applications.

脆弱水印是检测图像篡改的一种有效技术，但目前基于分块的脆弱水印方法往往存在水印图像质量有限、误报率高的问题。为了克服这些局限性，本文提出了一种基于改进的最低有效位替换（least significant bit substitution， ILSBS）的增强半盲脆弱水印方法TLFW。首先，将原始图像划分为互不重叠的块，在块内采用基于规则的ILSBS方法动态调整水印嵌入位，有效减小像素失真；这大大提高了水印图像的视觉质量。接下来，引入了篡改定位优化（TLO）策略，通过在参考点周围改进检测结果，大大减少误报。大量的实验结果表明，TLFW方法将水印图像的峰值信噪比（PSNR）从44 dB提高到46 dB，在多种攻击场景中持续降低假阳性率（FPR），同时在文本添加攻击中实现45%的FPR降低和52%的篡改检测率（TDR）增益，并且与现有方法相比显著降低了计算成本。所提出的TLFW方案兼容灰度图像和彩色图像，并且在水印提取时不需要原始图像，非常适合实际图像认证应用。

{"title":"Improved LSB substitution based semi-blind fragile watermarking for high-accuracy tamper localization","authors":"Pengfei Li , Weijia Cao , Tao Yu","doi":"10.1016/j.jisa.2025.104330","DOIUrl":"10.1016/j.jisa.2025.104330","url":null,"abstract":"<div><div>Fragile watermarking is an effective technique for detecting image tampering, yet current block-based fragile watermarking methods often struggle with limited watermarked image quality and high false-positive rates. To overcome these limitations, this paper proposes an enhanced semi-blind fragile watermarking method based on an improved least significant bit (LSB) substitution (ILSBS), named TLFW. First, the original image is divided into non-overlapping blocks, within which a novel rule-based ILSBS approach dynamically adjusts the watermark embedding bits to minimize pixel distortion effectively. This significantly enhances the visual quality of the watermarked image. Next, a tamper localization optimization (TLO) strategy is introduced to substantially reduce false positives by refining detection results around reference points. Extensive experimental results demonstrate that the TLFW approach improves the peak signal-to-noise ratio (PSNR) of watermarked images from 44 dB to 46 dB, consistently reduces the false positive rate (FPR) across multiple attack scenarios while achieving 45 % FPR reduction and 52 % tamper detection rate (TDR) gain specifically for the text-addition attack, and lowers computational costs significantly compared to existing methods. The proposed TLFW scheme is compatible with both grayscale and color images and does not require the original image during watermark extraction, making it highly suitable for practical image authentication applications.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104330"},"PeriodicalIF":3.7,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738764","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Software supply chain: A taxonomy of attacks, mitigations and risk assessment strategies 软件供应链：攻击、缓解和风险评估策略的分类

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-11 DOI: 10.1016/j.jisa.2025.104324

Betul Gokkaya, Leonardo Aniello, Basel Halak

The software product is a source of cyber-attacks that target organizations by using their software supply chain (SSC) as a distribution vector. As the reliance of software projects on open-source or proprietary modules is increasing drastically, SSC is becoming more and more critical and, therefore, has attracted the interest of cyber attackers. While existing studies primarily focus on software supply chain attacks’ prevention and detection methods, there is a need for a broad overview of attacks and comprehensive risk assessment for software supply chain security. This study conducts a systematic literature review to fill this gap. By analyzing 96 papers published between 2015-2023, we identified 19 distinct SSC attacks, including 6 novel attacks highlighted in recent studies. Additionally, we developed 25 specific security controls and established a precisely mapped taxonomy that transparently links each control to one or more specific attacks. By establishing this relationship, we demonstrate how SSC security controls are strategically designed to counteract specific attack vectors. Furthermore, we emphasize the role of risk assessment as a foundational step in understanding and prioritizing these vulnerabilities. This study introduces a risk assessment methodology tailored to software supply chain environments, focusing on identifying vulnerabilities in software components, dependencies, and suppliers. The proposed methodology enables organizations to systematically prioritize threats and implement appropriate mitigation strategies.

软件产品是网络攻击的一个来源，通过将组织的软件供应链（SSC）作为分发载体来攻击组织。随着软件项目对开源或专有模块的依赖急剧增加，SSC变得越来越重要，因此引起了网络攻击者的兴趣。现有的研究主要集中在软件供应链攻击的预防和检测方法上，需要对软件供应链安全的攻击进行广泛的概述和全面的风险评估。本研究通过系统的文献综述来填补这一空白。通过分析2015-2023年间发表的96篇论文，我们确定了19种不同的SSC攻击，包括最近研究中突出的6种新型攻击。此外，我们开发了25个特定的安全控件，并建立了一个精确映射的分类法，将每个控件透明地链接到一个或多个特定的攻击。通过建立这种关系，我们演示了如何战略性地设计SSC安全控制来对抗特定的攻击向量。此外，我们强调风险评估作为理解和优先考虑这些脆弱性的基础步骤的作用。本研究介绍了一种针对软件供应链环境的风险评估方法，着重于识别软件组件、依赖关系和供应商中的漏洞。拟议的方法使各组织能够系统地确定威胁的优先次序并实施适当的缓解战略。

{"title":"Software supply chain: A taxonomy of attacks, mitigations and risk assessment strategies","authors":"Betul Gokkaya, Leonardo Aniello, Basel Halak","doi":"10.1016/j.jisa.2025.104324","DOIUrl":"10.1016/j.jisa.2025.104324","url":null,"abstract":"<div><div>The software product is a source of cyber-attacks that target organizations by using their software supply chain (SSC) as a distribution vector. As the reliance of software projects on open-source or proprietary modules is increasing drastically, SSC is becoming more and more critical and, therefore, has attracted the interest of cyber attackers. While existing studies primarily focus on software supply chain attacks’ prevention and detection methods, there is a need for a broad overview of attacks and comprehensive risk assessment for software supply chain security. This study conducts a systematic literature review to fill this gap. By analyzing 96 papers published between 2015-2023, we identified 19 distinct SSC attacks, including 6 novel attacks highlighted in recent studies. Additionally, we developed 25 specific security controls and established a precisely mapped taxonomy that transparently links each control to one or more specific attacks. By establishing this relationship, we demonstrate how SSC security controls are strategically designed to counteract specific attack vectors. Furthermore, we emphasize the role of risk assessment as a foundational step in understanding and prioritizing these vulnerabilities. This study introduces a risk assessment methodology tailored to software supply chain environments, focusing on identifying vulnerabilities in software components, dependencies, and suppliers. The proposed methodology enables organizations to systematically prioritize threats and implement appropriate mitigation strategies.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104324"},"PeriodicalIF":3.7,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738765","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Anomaly detection for blockchain nodes based on eBPF and fine-tuning large language model 基于eBPF和大语言模型微调的区块链节点异常检测

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-10 DOI: 10.1016/j.jisa.2025.104329

Jincheng Su , Zhide Chen , Kexin Zhu , Chen Feng

While blockchain technology is widely used across various fields, it faces growing security challenges. Traditional blockchain anomaly detection methods, such as log analysis and fixed pattern recognition, struggle to handle complex and dynamic attack techniques. This paper proposes the Blockchain Live Anomaly Detection with eBPF and LLMs (BLAD-eLLM) scheme, which combines the efficient data capture capabilities of extended Berkeley Packet Filter (eBPF) technology for kernel-level security monitoring with the deep text understanding and semantic analysis power of large language models (LLMs) to enhance the network layer security of blockchain nodes. The proposed approach analyzes blockchain network activities comprehensively, aiming for accurate identification of potential anomalous behaviors. Furthermore, a knowledge-enhanced reasoning framework is developed, integrating Retrieval-Augmented Generation (RAG) for contextual blockchain threat intelligence and Chain-of-Thought (COT) prompting for multi-step attack analysis while employing Weight-Decomposed Low-Rank Adaptation (DoRA) based prompt fine-tuning to optimize the LLMs’ domain-specific adaptability and detection accuracy. Experimental results demonstrate that the proposed scheme significantly improves blockchain anomaly detection accuracy while maintaining minimal impact on system performance, ensuring the stability and security of the blockchain system.

区块链技术在各个领域得到广泛应用的同时，也面临着越来越多的安全挑战。传统的区块链异常检测方法，如日志分析和固定模式识别，难以处理复杂的动态攻击技术。本文提出了区块链实时异常检测与eBPF和llm （blade - ellm）方案，该方案将扩展伯克利包过滤（eBPF）技术用于内核级安全监控的高效数据捕获能力与大型语言模型（llm）的深度文本理解和语义分析能力相结合，以增强区块链节点的网络层安全性。该方法全面分析区块链网络活动，旨在准确识别潜在的异常行为。此外，开发了一个知识增强推理框架，集成了用于上下文区块链威胁情报的检索增强生成（RAG）和用于多步攻击分析的思维链（COT）提示，同时采用基于权重分解的低秩自适应（DoRA）提示微调来优化llm的特定领域适应性和检测精度。实验结果表明，该方案在保证区块链系统稳定性和安全性的同时，显著提高了区块链异常检测精度。

{"title":"Anomaly detection for blockchain nodes based on eBPF and fine-tuning large language model","authors":"Jincheng Su , Zhide Chen , Kexin Zhu , Chen Feng","doi":"10.1016/j.jisa.2025.104329","DOIUrl":"10.1016/j.jisa.2025.104329","url":null,"abstract":"<div><div>While blockchain technology is widely used across various fields, it faces growing security challenges. Traditional blockchain anomaly detection methods, such as log analysis and fixed pattern recognition, struggle to handle complex and dynamic attack techniques. This paper proposes the Blockchain Live Anomaly Detection with eBPF and LLMs (BLAD-eLLM) scheme, which combines the efficient data capture capabilities of extended Berkeley Packet Filter (eBPF) technology for kernel-level security monitoring with the deep text understanding and semantic analysis power of large language models (LLMs) to enhance the network layer security of blockchain nodes. The proposed approach analyzes blockchain network activities comprehensively, aiming for accurate identification of potential anomalous behaviors. Furthermore, a knowledge-enhanced reasoning framework is developed, integrating Retrieval-Augmented Generation (RAG) for contextual blockchain threat intelligence and Chain-of-Thought (COT) prompting for multi-step attack analysis while employing Weight-Decomposed Low-Rank Adaptation (DoRA) based prompt fine-tuning to optimize the LLMs’ domain-specific adaptability and detection accuracy. Experimental results demonstrate that the proposed scheme significantly improves blockchain anomaly detection accuracy while maintaining minimal impact on system performance, ensuring the stability and security of the blockchain system.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104329"},"PeriodicalIF":3.7,"publicationDate":"2025-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Security analysis of a novel image encryption algorithm based on 3D chaos 一种基于三维混沌的图像加密算法的安全性分析

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-09 DOI: 10.1016/j.jisa.2025.104332

Rong Zhou, Simin Yu

The paper analyzes the security of a novel encryption algorithm based on a 3D chaos from three aspects. The algorithm uses a typical two-round permutation-diffusion structure and consists of three parts: key selection, permutation, and diffusion with feedback. The keys consist of three initial values, three control parameters, one interference parameter, and two correlation coefficients of 3D chaos. The algorithm has passed the tests for histogram, correlation, difference analysis, and trend fluctuation analysis, and its security has been verified. However, after analysis in this paper, it has three major vulnerabilities: first, the algorithm has an equivalent key because the original key is irrelevant to plaintext. Second, diffusion is a linear operation, it results in the separability between the pixel values of the plaintext and the keys that form the pixel values of the ciphertext. Hence, the plaintext can be reconstructed according to the characteristics of ciphertext. Third, the feedback in diffusion makes ciphertext have a strong sensitivity to plaintext, but the sensitivity of plaintext to ciphertext is low and there is no avalanche effect. Therefore, it cannot resist differential attack. This paper analyzes the algorithm from three vulnerabilities, respectively, and proposes three decoding methods. Meanwhile, targeted measures are discussed to resist common attacks. Finally, the three methods are extended to cryptanalysis on general encryption algorithm with similar vulnerabilities. Theoretical analysis and experimental results show that these three decoding methods are effective.

本文从三个方面分析了一种基于三维混沌的新型加密算法的安全性。该算法采用典型的两轮置换扩散结构，由键选择、置换和带反馈的扩散三部分组成。键由3个初始值、3个控制参数、1个干涉参数和2个三维混沌相关系数组成。该算法通过了直方图、相关性、差异分析、趋势波动分析等测试，验证了算法的安全性。然而，经过本文的分析，该算法存在三大漏洞：首先，由于原密钥与明文无关，该算法具有等效密钥。其次，扩散是一种线性操作，它导致了明文的像素值与构成密文像素值的密钥之间的可分离性。因此，可以根据密文的特点对明文进行重构。第三，扩散中的反馈使得密文对明文具有较强的敏感性，但明文对密文的敏感性较低，不存在雪崩效应。因此，它无法抵抗差分攻击。本文分别从三个漏洞对该算法进行了分析，并提出了三种解码方法。同时，讨论了针对常见攻击的针对性措施。最后，将这三种方法推广到具有类似漏洞的通用加密算法的密码分析中。理论分析和实验结果表明，这三种译码方法都是有效的。

{"title":"Security analysis of a novel image encryption algorithm based on 3D chaos","authors":"Rong Zhou, Simin Yu","doi":"10.1016/j.jisa.2025.104332","DOIUrl":"10.1016/j.jisa.2025.104332","url":null,"abstract":"<div><div>The paper analyzes the security of a novel encryption algorithm based on a 3D chaos from three aspects. The algorithm uses a typical two-round permutation-diffusion structure and consists of three parts: key selection, permutation, and diffusion with feedback. The keys consist of three initial values, three control parameters, one interference parameter, and two correlation coefficients of 3D chaos. The algorithm has passed the tests for histogram, correlation, difference analysis, and trend fluctuation analysis, and its security has been verified. However, after analysis in this paper, it has three major vulnerabilities: first, the algorithm has an equivalent key because the original key is irrelevant to plaintext. Second, diffusion is a linear operation, it results in the separability between the pixel values of the plaintext and the keys that form the pixel values of the ciphertext. Hence, the plaintext can be reconstructed according to the characteristics of ciphertext. Third, the feedback in diffusion makes ciphertext have a strong sensitivity to plaintext, but the sensitivity of plaintext to ciphertext is low and there is no avalanche effect. Therefore, it cannot resist differential attack. This paper analyzes the algorithm from three vulnerabilities, respectively, and proposes three decoding methods. Meanwhile, targeted measures are discussed to resist common attacks. Finally, the three methods are extended to cryptanalysis on general encryption algorithm with similar vulnerabilities. Theoretical analysis and experimental results show that these three decoding methods are effective.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"97 ","pages":"Article 104332"},"PeriodicalIF":3.7,"publicationDate":"2025-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738762","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

BISE: Enhance data sharing security through consortium blockchain and IPFS BISE：通过联盟区块链和IPFS增强数据共享安全性

IF 3.7 2区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Journal of Information Security and Applications

Pub Date : 2025-12-05 DOI: 10.1016/j.jisa.2025.104320

Mingxuan Chen , Puhe Hao , Weizhi Meng , Yasen Aizezi , Guozi Sun

Data sharing is pivotal in sectors such as healthcare, finance, and social networking. Encrypting sensitive data, while essential for privacy protection, introduces complexity to data sharing and poses privacy risks when leveraging cloud servers. Blockchain-based searchable encryption offers a balance between privacy preservation and data availability; however, user anonymity remains a significant concern. Traditional storage systems, which rely on centralized servers, limit data stability and scalability. To address these challenges, we have introduced BISE, a solution that leverages the power of blockchain to achieve data integrity, using searchable encryption for secure searches and IPFS for decentralized storage. Constructed on Hyperledger Fabric and IPFS, our system demonstrates efficiency through simulations. This integrated approach ensures data privacy, integrity, and availability, with efficient updates and queries, making it a robust solution for sensitive data sharing in various domains.

数据共享在医疗保健、金融和社交网络等领域至关重要。对敏感数据进行加密虽然对隐私保护至关重要，但会给数据共享带来复杂性，并在利用云服务器时带来隐私风险。基于区块链的可搜索加密在隐私保护和数据可用性之间提供了平衡；然而，用户匿名仍然是一个重大问题。传统的存储系统依赖于集中式服务器，限制了数据的稳定性和可扩展性。为了应对这些挑战，我们引入了BISE，这是一种利用区块链功能实现数据完整性的解决方案，使用可搜索加密进行安全搜索，使用IPFS进行分散存储。本系统基于Hyperledger Fabric和IPFS架构，通过仿真验证了系统的有效性。这种集成的方法确保数据隐私、完整性和可用性，并具有高效的更新和查询，使其成为各种领域中敏感数据共享的健壮解决方案。

引用次数: 0