- User Authentication and Security Systems
- Speech and Audio Processing
- Advanced Malware Detection Techniques
- Speech Recognition and Synthesis
- Digital Media Forensic Detection
- Physical Unclonable Functions (PUFs) and Hardware Security
- Adversarial Robustness in Machine Learning
- Network Security and Intrusion Detection
- Electrostatic Discharge in Electronics
- Cryptographic Implementations and Security
- Caching and Content Delivery
- Cooperative Communication and Network Coding
- Mobile Ad Hoc Networks
- Peer-to-Peer Network Technologies
- Smart Grid Security and Resilience
- Face recognition and analysis
- Biometric Identification and Security
- Music and Audio Processing
- Mobile Agent-Based Network Management
- Network Traffic and Congestion Control
- Opportunistic and Delay-Tolerant Networks
- Electromagnetic Compatibility and Noise Suppression
- Power Systems and Technologies
- Advanced Optical Sensing Technologies
- Full-Duplex Wireless Communications
Zhejiang University
2011-2025
University of British Columbia
2025
Suzhou University of Science and Technology
2025
Huawei Technologies (United Kingdom)
2023
Northwestern University
2006-2020
Hefei University of Technology
2020
Ministry of Education of the People's Republic of China
2020
Anhui Xinhua University
2018
University of Colorado Boulder
2013-2014
University of Maryland, College Park
2012-2013
Speech recognition (SR) systems such as Siri or Google Now have become an increasingly popular human-computer interaction method, and turned various into voice controllable (VCS). Prior work on attacking VCS shows that the hidden commands are incomprehensible to people can control systems. Hidden commands, though "hidden", nonetheless audible. In this work, we design a totally inaudible attack, DolphinAttack, modulates ultrasonic carriers (e.g., f > 20 kHz) achieve inaudibility. By...
Zero-day polymorphic worms pose a serious threat to the security of Internet infrastructures. Given their rapid propagation, it is crucial detect them at edge networks and automatically generate signatures in early stages infection. Most existing approaches for automatic signature generation need host information are thus not applicable deployment on high-speed network links. In this paper, we propose Hamsa, network-based automated system which fast, noise-tolerant attack-resilient....
Autonomous vehicles rely on sensors to measure road condition and make driving decisions, their safety relies heavily the reliability of these sensors. Out all obstacle detection sensors, ultrasonic have largest market share are expected be increasingly installed automobiles. Such discover obstacles by emitting ultrasounds analyzing reflections. By exploiting built-in vulnerabilities we designed random spoofing, adaptive jamming attacks managed trick a vehicle stop when it should keep...
Voice assistants (VAs) such as Siri and Google Now have become an increasingly popular human-machine interaction method made various systems voice controllable. Prior work on attacking shows that the hidden commands are incomprehensible to people can control VAs. Hidden commands, though `hidden', nonetheless audible. In this work, we design a completely inaudible attack, DolphinAttack, modulates ultrasonic carriers achieve inaudibility. By leveraging nonlinearity of microphone circuits,...
Recent hardware advances have led to the development and consumerization of mobile devices, which mainly include smartphones various wearable devices. To protect privacy users, user authentication mechanisms been proposed. In particular, biometrics has widely used for multi-factor authentication. However, biometrics-based usually require costly sensors deployed on rely explicit input Internet connection performing this article, we propose a system, called RISKCOG, can authenticate ownership...
Over the last six years, several papers demonstrated how intentional analog interference based on acoustics, RF, lasers, and other physical modalities could induce faults, influence, or even control output of sensors. Damage to availability integrity sensor carries significant risks safety-critical systems that make automated decisions trusted measurement. Established signal processing models use transfer functions express reliability dependability characteristics sensors, but existing do...
Verifying the identity of voice inputs is important as voices are increasingly used for sensitive operations. Traditional methods focus on differentiating individuals via spectrographic features (e.g., voiceprint), yet cannot cope with spoofing attacks, whereby a malicious attacker synthesizes almost same voiceprint victim or simply replays it. This paper proposes CaField, text-independent speaker verification method to detect loudspeaker-based attacks goal achieving two seemingly...
Autonomous vehicles and robots increasingly exploit LiDAR-based 3D object detection systems to detect obstacles in environment. Correct classification are important ensure safe driving. Though existing work has demonstrated the feasibility of manipulating point clouds spoof detectors, most attempts conducted digitally. In this paper, we investigate possibility physically fooling by injecting adversarial using lasers. First, develop a laser transceiver that can inject up 4200 points, which is...
Recently, there has been an increasing deployment of content distribution networks (CDNs) that offer hosting services to Web providers. In this paper, we first compare the uncooperative pulling contents used by commercial CDNs with cooperative pushing. Our results show latter can achieve comparable users' perceived performance only 4%-5% replication and update traffic compared former scheme. Therefore, explore how efficiently push CDN nodes. Using trace-driven simulation, replicating in...
<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> This letter addresses the security issues of multimedia encryption schemes using multiple Huffman table (MHT). A known-plaintext attack is presented to show that MHTs used for should be carefully selected avoid weak keys problem. We then propose chosen-plaintext attacks on basic MHT algorithm as well enhanced scheme with random bit insertion. In addition, we suggest two empirical criteria...
Autonomous vehicles increasingly exploit computer-vision-based object detection systems to perceive environments and make critical driving decisions. To increase the quality of images, image stabilizers with inertial sensors are added alleviate blurring caused by camera jitters. However, such a trend opens new attack surface. This paper identifies system-level vulnerability resulting from combination emerging stabilizer hardware susceptible acoustic manipulation algorithms subject...
As renewable energy sources (RES) continue to expand and the use of power inverters has surged, have become crucial for converting direct current (DC) from RES into alternating (AC) grid, their security is vital maintaining stable grid operations. This paper investigates vulnerabilities photovoltaic (PV) inverters, specifically focusing on internal sensors, which are critical reliable conversion. It found that both voltage sensors susceptible intentional electromagnetic interference (IEMI)...
We put forward a physical model of uniformly accelerated Unruh-DeWitt battery and use quantum work extraction as probe to witness the thermal nature Unruh effect in high dimensional Minkowski spacetime. By means open system approach, we investigate maximal amount with respect acceleration-induced temperature, spacetime dimensionality, field mass. It has been found that steady asymptotic condition is just determined by temperature arbitrary spacetimes. The behavior can demonstrate global...
Cooperative diversity is an effective technique to combat multipath fading. When this applied ad hoc networks, it a key issue design distributed and power-efficient relay selection strategies. In paper, we propose the power-aware (PARS) strategies maximize network lifetime. The PARS are composed of optimal power allocation(OPA) three criteria. OPA aims at minimizing overall transmit power, while criteria select that best help extend lifetime, based on both results residual levels nodes. We...
Botnets dominate today's attack landscape. In this work, we investigate ways to analyze collections of malicious probing traffic in order understand the significance large-scale “botnet probes.” such events, an entire collection remote hosts together probes address space monitored by a sensor some sort coordinated fashion. Our goal is develop methodologies which sites receiving can infer-using purely local observation-information about activity: What scanning strategies does employ? Is that...
Text-to-Speech (TTS) and Voice Conversion (VC) models have exhibited remarkable performance in generating realistic natural audio. However, their dark side, audio deepfake poses a significant threat to both society individuals. Existing countermeasures largely focus on determining the genuineness of speech based complete original recordings, which however often contain private content. This oversight may refrain detection from many applications, particularly scenarios involving sensitive...
Relayed transmission is an efficient way to combat wireless channel fading and increase system throughput. By allocating the overall power transmitter relay optimally, performance can be further improved. This paper discusses optimal allocation in frequency-selective non-regenerative relaying channels, where OFDM employed. Firstly, between a cooperative subcarrier pair analyzed equivalent gain model for pairs developed. Based on this, sub-optimal selection method given by applying Hungarian...
Proxy caching servers are widely deployed in today's Internet. While cooperation among proxy caches can significantly improve a network's resilience to denial-of-service (DoS) attacks, lack of transform such into viable DoS targets. In this paper, we investigate class pollution attacks that aim degrade proxy's capabilities, either by ruining the cache file locality, or inducing false locality. Using simulations, propose and evaluate effects both Web peer-to- peer (p2p) scenarios, reveal...
Packet forwarding prioritization (PFP) in routers is one of the mechanisms commonly available to network operators. PFP can have a significant impact on accuracy measurements, performance applications and effectiveness troubleshooting procedures. Despite its potential impacts, no information settings readily end users. In this paper, we present an end-to-end approach for inference associated tool, POPI. This <i xmlns:mml="http://www.w3.org/1998/Math/MathML"...