The topology of a network is fundamental for building network infrastructure functionalities. In many scenarios, enterprise networks may have no desire to disclose their topology information. In this paper, we aim at preventing attacks that use adversarial, active end-to-end topology inference to obtain the topology information of a target network. To this end, we propose a Proactive Topology Obfuscation (ProTO) system that adopts a detect-then-obfuscate framework: (i) a lightweight probing behavior identification mechanism based on machine learning is designed to detect any probing behavior, and then (ii) a topology obfuscation design is developed to proactively delay all identified probe packets in a way such that the attacker will obtain a structurally accurate yet fake network topology based on the measurements of these delayed probe packets, therefore deceiving the attacker and decreasing its appetency for future inference. We evaluate ProTO under different evaluation scenarios. Experimental results show that ProTO is able to (i) achieve a detection rate of 99.9% with a false alarm of 3%, (ii) effectively disrupt adversarial topology inference and lead to the topology inferred by the attacker close to a fake topology, and (iii) result in an overall network delay performance degradation of 1.3% - 2.0%.

The convergence of ubiquitous connectivity, large-scale data generation, and rapid advancements in machine learning is transforming the field of cybersecurity. The widespread adoption of interconnected systems including Internet of Things devices, mobile platforms, and cloud infrastructures has introduced new attack surfaces and significantly increased the complexity of securing digital environments. Concurrently, these technologies have enabled the development of intelligent, data-driven defense strategies. Achieving effective protection in these settings requires not only applying machine learning to detect and prevent threats but also recognizing that such models can themselves become targets of adversarial manipulation. This survey presents a comprehensive analysis of recent progress at the intersection of machine learning and cybersecurity. It explores defensive applications such as malware detection, network traffic classification, and anomaly detection, as well as offensive strategies including adversarial evasion, poisoning, and backdoor attacks. Particular attention is paid to adversarial machine learning, highlighting the increasing sophistication of attacks that exploit model vulnerabilities and the corresponding evolution of defense mechanisms. Beyond synthesizing current research, the survey also identifies key open challenges and emerging research directions. This survey provides a comprehensive and accessible reference for researchers and practitioners aiming to understand and advance the secure application of machine learning across diverse cybersecurity domains.

The Internet of Vehicles (IoV) marks a revolutionary leap in transportation, integrating vehicles with the Internet to enhance convenience, intelligence, and efficiency. As IoV applications continue to expand, a spectrum of challenges and untapped opportunities emerge. Among these, the security of IoV stands out as a critical issue in transportation systems, given its direct impact on road safety. To this end, this survey conducts an in-depth analysis of IoV security challenges and introduces a distinctive approach by systematically categorizing threats into inside-vehicle and outside-vehicle domains, providing a comprehensive understanding of the full spectrum of risks. Meanwhile, we examine defense mechanisms designed to counter these threats and explore proactive strategies to enhance IoV security. In addition, this study explores emerging protection techniques, such as AI-driven intrusion detection, blockchain-based trust management, and 5G-enabled secure routing, demonstrating how these technologies can be effectively integrated to safeguard IoV systems. Furthermore, this survey provides actionable recommendations and forward-looking research directions to guide the development of real-world implementations, standardized procedures, and regulatory frameworks, benefiting stakeholders, policymakers, and researchers. As IoV continues its rapid evolution, these insights offer a comprehensive roadmap to strengthen IoV security and contribute to a safer, more resilient intelligent transportation ecosystem.

Wireless networks nowadays are ubiquitous in our daily life. Due to the open channel nature, wireless networks are vulnerable to eavesdropping attacks. Though wireless conversation can be encrypted against eavesdropping, attackers can still infer a user’s activities via traffic analysis on encrypted data. Nevertheless, previous inference methods usually have the limitation that they can only achieve a relatively high accuracy in a specific domain. In this paper, we propose a smart spying strategy that can infer a user’s activities of multiple domains with a higher accuracy. We also develop a prototype tool on top of this strategy to conduct experiments. The evaluation results show our strategy works effectively in activity inference on encrypted data, with an accuracy rate as high as 99.17%.

The evolution of the next-generation network has indeed facilitated seamless communication and coexistence of diverse devices with varying capabilities and connectivity requirements in modern communication systems. However, it also incurs a wide range of network attacks (e.g., DDoS, U2R, R2L, and Reconnaissance attacks). Among them, reconnaissance attacks seem receiving less attention as they are passively launched and do not directly impact the network. Nevertheless, reconnaissance attacks could serve as a critical preliminary step of advanced attacks and should never be left behind. In this work, we develop a novel neural learning framework that incorporates feature selection and model training together to achieve a better performance in reconnaissance attack detection. In particular, our framework contains two key techniques: 1) Self-adaptive feature selection integrates the attention and dropping mechanisms to facilitate dynamic feature set adaption. 2) Lightweight attention-based LSTM for feature learning. It allows us to attend to relevant packets within each traffic flow and extract the temporal dependency of reconnaissance behaviors for better attack identification. In the experiment, we build a real-world network testbed to validate our design. The results show that the detection model can achieve a detection rate of up to 99.88% with the proposed neural learning detector.