arXiv subjects

Pu Wang

Publications and source records attributed to Pu Wang.

At least 109 records · Page 6Linked to original sources

arXiv2021

MutualNet: Adaptive ConvNet via Mutual Learning from Different Model Configurations

Most existing deep neural networks are static, which means they can only do inference at a fixed complexity. But the resource budget can vary substantially across different devices. Even on a single device, the affordable budget can change with different scenarios, and repeatedly training networks for each required budget would be incredibly expensive. Therefore, in this work, we propose a general method called MutualNet to train a single network that can run at a diverse set of resource constraints. Our method trains a cohort of model configurations with various network widths and input resolutions. This mutual learning scheme not only allows the model to run at different width-resolution configurations but also transfers the unique knowledge among these configurations, helping the model to learn stronger representations overall. MutualNet is a general training methodology that can be applied to various network structures (e.g., 2D networks: MobileNets, ResNet, 3D networks: SlowFast, X3D) and various tasks (e.g., image classification, object detection, segmentation, and action recognition), and is demonstrated to achieve consistent improvements on a variety of datasets. Since we only train the model once, it also greatly reduces the training cost compared to independently training several models. Surprisingly, MutualNet can also be used to significantly boost the performance of a single network, if dynamic resource constraint is not a concern. In summary, MutualNet is a unified method for both static and adaptive, 2D and 3D networks. Codes and pre-trained models are available at \url{https://github.com/taoyang1122/MutualNet}.

cs.CV↗

arXiv2021

Discrete-modulation continuous-variable quantum key distribution with high key rate

Discrete-modulation continuous-variable quantum key distribution has the potential for large-scale deployment in the secure quantum communication networks due to low implementation complexity and compatibility with the current telecom systems. The security proof for four coherent states phase-shift keying (4-PSK) protocol has recently been established by applying numerical methods. However, the achievable key rate is relatively low compared with the optimal Gaussian modulation scheme. To enhance the key rate of discrete-modulation protocol, we first show that 8-PSK increases the key rate by about 60\% in comparison to 4-PSK, whereas the key rate has no significant improvement from 8-PSK to 12-PSK. We then expand the 12-PSK to two-ring constellation structure with four states in the inner ring and eight states in the outer ring, which significantly improves the key rate to be 2.4 times of that of 4-PSK. The key rate of the two-ring constellation structure can reach 70\% of the key rate achieved by Gaussian modulation in long distance transmissions, making this protocol an attractive alternative for high-rate and low-cost application in secure quantum communication networks.

quant-ph↗

arXiv2021

Sim-to-Real Transfer in Multi-agent Reinforcement Networking for Federated Edge Computing

Federated Learning (FL) over wireless multi-hop edge computing networks, i.e., multi-hop FL, is a cost-effective distributed on-device deep learning paradigm. This paper presents FedEdge simulator, a high-fidelity Linux-based simulator, which enables fast prototyping, sim-to-real code, and knowledge transfer for multi-hop FL systems. FedEdge simulator is built on top of the hardware-oriented FedEdge experimental framework with a new extension of the realistic physical layer emulator. This emulator exploits trace-based channel modeling and dynamic link scheduling to minimize the reality gap between the simulator and the physical testbed. Our initial experiments demonstrate the high fidelity of the FedEdge simulator and its superior performance on sim-to-real knowledge transfer in reinforcement learning-optimized multi-hop FL.

cs.NI↗

arXiv2021

A Study into Pre-training Strategies for Spoken Language Understanding on Dysarthric Speech

End-to-end (E2E) spoken language understanding (SLU) systems avoid an intermediate textual representation by mapping speech directly into intents with slot values. This approach requires considerable domain-specific training data. In low-resource scenarios this is a major concern, e.g., in the present study dealing with SLU for dysarthric speech. Pretraining part of the SLU model for automatic speech recognition targets helps but no research has shown to which extent SLU on dysarthric speech benefits from knowledge transferred from other dysarthric speech tasks. This paper investigates the efficiency of pre-training strategies for SLU tasks on dysarthric speech. The designed SLU system consists of a TDNN acoustic model for feature encoding and a capsule network for intent and slot decoding. The acoustic model is pre-trained in two stages: initialization with a corpus of normal speech and finetuning on a mixture of dysarthric and normal speech. By introducing the intelligibility score as a metric of the impairment severity, this paper quantitatively analyzes the relation between generalization and pathology severity for dysarthric speech.

eess.AS↗

arXiv2021

Pre-training for low resource speech-to-intent applications

Designing a speech-to-intent (S2I) agent which maps the users' spoken commands to the agents' desired task actions can be challenging due to the diverse grammatical and lexical preference of different users. As a remedy, we discuss a user-taught S2I system in this paper. The user-taught system learns from scratch from the users' spoken input with action demonstration, which ensure it is fully matched to the users' way of formulating intents and their articulation habits. The main issue is the scarce training data due to the user effort involved. Existing state-of-art approaches in this setting are based on non-negative matrix factorization (NMF) and capsule networks. In this paper we combine the encoder of an end-to-end ASR system with the prior NMF/capsule network-based user-taught decoder, and investigate whether pre-training methodology can reduce training data requirements for the NMF and capsule network. Experimental results show the pre-trained ASR-NMF framework significantly outperforms other models, and also, we discuss limitations of pre-training with different types of command-and-control(C&C) applications.

eess.AS↗

arXiv2020

A light transformer for speech-to-intent applications

Spoken language understanding (SLU) systems can make life more agreeable, safer (e.g. in a car) or can increase the independence of physically challenged users. However, due to the many sources of variation in speech, a well-trained system is hard to transfer to other conditions like a different language or to speech impaired users. A remedy is to design a user-taught SLU system that can learn fully from scratch from users' demonstrations, which in turn requires that the system's model quickly converges after only a few training samples. In this paper, we propose a light transformer structure by using a simplified relative position encoding with the goal to reduce the model size and improve efficiency. The light transformer works as an alternative speech encoder for an existing user-taught multitask SLU system. Experimental results on three datasets with challenging speech conditions prove our approach outperforms the existed system and other state-of-art models with half of the original model size and training time.

eess.AS↗

arXiv2020

Who is in Control? Practical Physical Layer Attack and Defense for mmWave based Sensing in Autonomous Vehicles

With the wide bandwidths in millimeter wave (mmWave) frequency band that results in unprecedented accuracy, mmWave sensing has become vital for many applications, especially in autonomous vehicles (AVs). In addition, mmWave sensing has superior reliability compared to other sensing counterparts such as camera and LiDAR, which is essential for safety-critical driving. Therefore, it is critical to understand the security vulnerabilities and improve the security and reliability of mmWave sensing in AVs. To this end, we perform the end-to-end security analysis of a mmWave-based sensing system in AVs, by designing and implementing practical physical layer attack and defense strategies in a state-of-the-art mmWave testbed and an AV testbed in real-world settings. Various strategies are developed to take control of the victim AV by spoofing its mmWave sensing module, including adding fake obstacles at arbitrary locations and faking the locations of existing obstacles. Five real-world attack scenarios are constructed to spoof the victim AV and force it to make dangerous driving decisions leading to a fatal crash. Field experiments are conducted to study the impact of the various attack scenarios using a Lincoln MKZ-based AV testbed, which validate that the attacker can indeed assume control of the victim AV to compromise its security and safety. To defend the attacks, we design and implement a challenge-response authentication scheme and a RF fingerprinting scheme to reliably detect aforementioned spoofing attacks.

cs.CR↗

arXiv2020

AST-GCN: Attribute-Augmented Spatiotemporal Graph Convolutional Network for Traffic Forecasting

Traffic forecasting is a fundamental and challenging task in the field of intelligent transportation. Accurate forecasting not only depends on the historical traffic flow information but also needs to consider the influence of a variety of external factors, such as weather conditions and surrounding POI distribution. Recently, spatiotemporal models integrating graph convolutional networks and recurrent neural networks have become traffic forecasting research hotspots and have made significant progress. However, few works integrate external factors. Therefore, based on the assumption that introducing external factors can enhance the spatiotemporal accuracy in predicting traffic and improving interpretability, we propose an attribute-augmented spatiotemporal graph convolutional network (AST-GCN). We model the external factors as dynamic attributes and static attributes and design an attribute-augmented unit to encode and integrate those factors into the spatiotemporal graph convolution model. Experiments on real datasets show the effectiveness of considering external information on traffic forecasting tasks when compared to traditional traffic prediction methods. Moreover, under different attribute-augmented schemes and prediction horizon settings, the forecasting accuracy of the AST-GCN is higher than that of the baselines.

cs.LG↗

arXiv2020

Time-Spatial Serials Differences' Probability Distribution of Natural Dynamical Systems

The normal distribution is used as a unified probability distribution, however, our researcher found that it is not good agreed with the real-life dynamical system's data. We collected and analyzed representative naturally occurring data series (e.g., the earth environment, sunspots, brain waves, electrocardiograms, some cases are classic chaos systems and social activities). It is found that the probability density functions (PDFs) of first or higher order differences for these datasets are consistently fat-tailed bell-shaped curves, and their associated cumulative distribution functions (CDFs) are consistently S-shaped when compared to the near-straight line of the normal distribution CDF. It is proved that this profile is not because of numerical or measure error, and the t-distribution is a good approximation. This kind of PDF/CDF is a universal phenomenon for independent time and space series data, which will make researchers to reconsider some hypotheses about stochastic dynamical models such as Wiener process, and therefore merits investigation.

math.DS↗

arXiv2020

High-fidelity, low-latency polarization quantum state transmissions over a hollow-core conjoined-tube fibre at around 800 nm

The performances of optical fibre-based quantum information systems are limited by the intrinsic properties of silica glass materials, e.g. high latency, Rayleigh-scattering loss wavelength scaling law, and cross-coupling induced modal impurity. Hollow-core optical fibre (HCF) promises to unify air-borne light propagation and non-line-of-sight transmission, thus holding great potentials for versatile photonics-based quantum infor-mation applications. The early version of HCF based on photonic-bandgap guidance has not proven itself as a reliable quantum channel because of the poor modal purity in both spatial and polarization domains, as well as significant difficulty in fabrication when the wavelength shifts to the visible region. In this work, based on the polarization degree of freedom, we first, to the best of our knowledge, demonstrate high-fidelity (~0.98) single-photon transmission and distribution of entangled photons over a conjoined-tube hollow-core fibre (CTF) by using commercial silicon single-photon avalanche photodiodes. Our CTF realized the combined merits of low loss, high spatial mode purity, low polarization degradation, and low chromatic dispersion. We also demonstrate single-photon low latency (~99.96% speed of light in vacuum) transmission, thus paving the way for extensive uses of HCF links in versatile polarization-based quantum information processing.

physics.optics↗

arXiv2020

Rapid Determination of Antimicrobial Susceptibility by Stimulated Raman Scattering Imaging of D2O Metabolic Incorporation in a Single Bacterium

Rapid antimicrobial susceptibility testing (AST) is urgently needed for treating infections with correct antibiotics and slowing down the emergence of antibiotic-resistant bacteria. Here, we report a phenotypic platform that rapidly produces AST results by femtosecond stimulated Raman scattering imaging of deuterium oxide (D2O) metabolism. Metabolic incorporation of D2O into biomass in a single bacterium is probed in as short as 10 minutes after culture in 70% D2O medium, the fastest among current technologies. Single-cell metabolism inactivation concentration (SC-MIC) is obtained in less than 2.5 hours from colony to results. The SC-MIC results of 37 sets of samples, which include 8 major bacterial species and 14 different antibiotics often encountered in clinic, are validated by standard minimal inhibitory concentration blindly measured via broth microdilution. Towards clinical translation, SRS imaging of D2O metabolic incorporation and SC-MIC determination after 1-h antibiotics treatment and 30-minutes mixture of D2O and antibiotics incubation of bacteria in urine or whole blood is demonstrated.

physics.bio-ph↗

arXiv2019

Machine Learning-Based Delay-Aware UAV Detection and Operation Mode Identification over Encrypted Wi-Fi Traffic

The consumer UAV (unmanned aerial vehicle) market has grown significantly over the past few years. Despite its huge potential in spurring economic growth by supporting various applications, the increase of consumer UAVs poses potential risks to public security and personal privacy. To minimize the risks, efficiently detecting and identifying invading UAVs is in urgent need for both invasion detection and forensics purposes. Given the fact that consumer UAVs are usually used in a civilian environment, existing physical detection methods (such as radar, vision, and sound) may become ineffective in many scenarios. Aiming to complement the existing physical detection mechanisms, we propose a machine learning-based framework for fast UAV identification over encrypted Wi-Fi traffic. It is motivated by the observation that many consumer UAVs use Wi-Fi links for control and video streaming. The proposed framework extracts features derived only from packet size and inter-arrival time of encrypted Wi-Fi traffic, and can efficiently detect UAVs and identify their operation modes. In order to reduce the online identification time, our framework adopts a re-weighted $\ell_1$-norm regularization, which considers the number of samples and computation cost of different features. This framework jointly optimizes feature selection and prediction performance in a unified objective function. To tackle the packet inter-arrival time uncertainty when optimizing the trade-off between the detection accuracy and delay, we utilize Maximum Likelihood Estimation (MLE) method to estimate the packet inter-arrival time. We collect a large number of real-world Wi-Fi data traffic of eight types of consumer UAVs and conduct extensive evaluation on the performance of our proposed method.

cs.CR↗

arXiv2019

Underwater Cooperative MIMO Communications using Hybrid Acoustic and Magnetic Induction Technique

Future smart ocean applications require long distance and reliable communications to connect underwater sensors/robots with remote surface base stations. It is challenging to achieve such goal due to the harsh and dynamic underwater acoustic channel. While Multiple-Input and Multiple-Output (MIMO) technique can enhance reliability and transmission range, it is difficult to place multiple acoustic transducers on one single underwater device due to the large wavelength. Although the cooperative MIMO technique that let multiple underwater devices form a virtual MIMO system could solve the issue, it was impossible to synchronize the distributed underwater devices due to the extremely large and dynamic propagation delay of acoustic waves. To this end, this paper proposes an underwater cooperative MIMO communication mechanism, which is based on a hybrid acoustic and Magnetic Induction (MI) technique. The inter-node synchronization problem can be perfectly solved by using the MI technique so that the distributed acoustic transducers can cooperatively form narrow beams for long distance underwater communications. The synchronization time and errors are significantly reduced since MI has negligible signal propagation delays. To quantitatively analyze the improvement, the closed-form expressions of the synchronization error, signal-to-noise ratio (SNR), effective communication time, and throughput of the proposed system are rigorously derived. The proposed hybrid system is implemented in a software-defined testbed under the beamforming and space-time coding scheme. Through both numerical analysis and real-world experiments, the paper shows that the proposed hybrid cooperative MIMO mechanism achieves much lower bit error rate and synchronization error than the conventional acoustic systems.

eess.SP↗

arXiv2019

Conquering the Rayleigh scattering limit of silica glass fiber at visible wavelengths with a hollow-core fiber approach

The ultimate limit on fiber loss is set by the intrinsic Rayleigh scattering of silica glass material. Here, we challenge this limit in the visible region by using a hollow-core fiber approach. Two visible-guiding hollow-core conjoined-tube negative-curvature fibers are successfully fabricated and exhibit the overall losses of 3.8 dB/km at 680 nm and 4.9 dB/km at 558 nm respectively. The loss of the latter fiber surpasses the Rayleigh scattering limit of silica glass fiber in the green spectral region by 2 dB. Numerical simulation indicates that this loss level is still much higher than the fundamental surface scattering loss limit of hollow-core fiber.

physics.optics↗

arXiv2019

Physical Layer Key Generation in 5G Wireless Networks

The bloom of the fifth generation (5G) communication and beyond serves as a catalyst for physical layer key generation techniques. In 5G communications systems, many challenges in traditional physical layer key generation schemes, such as co-located eavesdroppers, the high bit disagreement ratio, and high temporal correlation, could be overcome. This paper lists the key-enabler techniques in 5G wireless networks, which offer opportunities to address existing issues in physical layer key generation. We survey the existing key generation methods and introduce possible solutions for the existing issues. The new solutions include applying the high signal directionality in beamforming to resist co-located eavesdroppers, utilizing the sparsity of millimeter wave (mmWave) channel to achieve a low bit disagreement ratio under low signal-to-noise-ratio (SNR), and exploiting hybrid precoding to reduce the temporal correlation among measured samples. Finally, the future trends of physical layer key generation in 5G and beyond communications are discussed.

eess.SP↗

arXiv2019

Measuring Road Network Topology Vulnerability by Ricci Curvature

Describing the basic properties of road network systems, such as their robustness, vulnerability, and reliability, has been a very important research topic in the field of urban transportation. Current research mainly uses several statistical indicators of complex networks to analyze the road network systems. However, these methods are essentially node-based. These node-based methods are more concerned with the number of connections between nodes, and lack of consideration for interactions. So, this leads to the well-known node paradox problem, and their ability of characterizing the local and intrinsic properties of a network is weak. From the perspective of network intrinsic geometry, this paper proposes a method for measuring road network vulnerability using a discrete Ricci curvature, which can identify the key sections of a road network and indicate its fragile elements. The results show that our method performs better than complex network statistics on measuring the vulnerability of a road network. Additionally, it can characterize the evolution of the road network vulnerability among different periods of time in the same city through our method. Finally, we compare our method with the previous method of centrality and show the different between them. This article provides a new perspective on a geometry to analyze the vulnerability of a road network and describes the inherent nature of the vulnerability of a road system from a new perspective. It also contributes to enriching the analytical methods of complex road networks.

cs.SI↗

arXiv2019

On Reliability of Underwater Magnetic Induction Communications with Tri-Axis Coils

Underwater magnetic induction communications (UWMICs) provide a low-power and high-throughput solution for autonomous underwater vehicles (AUVs), which are envisioned to explore and monitor the underwater environment. UWMIC with tri-axis coils increases the reliability of the wireless channel by exploring the coil orientation diversity. However, the UWMIC channel is different from typical fading channels and the mutual inductance information (MII) is not always available. It is not clear the performance of the tri-axis coil MIMO without MII. Also, its performances with multiple users have not been investigated. In this paper, we analyze the reliability and multiplexing gain of UWMICs with tri-axis coils by using coil selection. We optimally select the transmit and receive coils to reduce the computation complexity and power consumption and explore the diversity for multiple users. We find that without using all the coils and MII, we can still achieve reliability. Also, the multiplexing gain of UWMIC without MII is 5dB smaller than typical terrestrial fading channels. The results of this paper provide a more power-efficient way to use UWMICs with tri-axis coils.

eess.SP↗

arXiv2019

Physical Layer Identification based on Spatial-temporal Beam Features for Millimeter Wave Wireless Networks

With millimeter wave (mmWave) wireless communication envisioned to be the key enabler of next generation high data rate wireless networks, security is of paramount importance. While conventional security measures in wireless networks operate at a higher layer of the protocol stack, physical layer security utilizes unique device dependent hardware features to identify and authenticate legitimate devices. In this work, we identify that the manufacturing tolerances in the antenna arrays used in mmWave devices contribute to a beam pattern that is unique to each device, and to that end we propose a novel device fingerprinting scheme based on the unique beam patterns used by the mmWave devices. Specifically, we propose a fingerprinting scheme with multiple access points (APs) to take advantage of the rich spatial-temporal information of the beam pattern. We perform comprehensive experiments with commercial off-the-shelf mmWave devices to validate the reliability performance of our proposed method under various scenarios. We also compare our beam pattern feature with a conventional physical layer feature namely power spectral density feature (PSD). To that end, we implement PSD feature based fingerprinting for mmWave devices. We show that the proposed multiple APs scheme is able to achieve over 99% identification accuracy for stationary LOS and NLOS scenarios and significantly outperform the PSD based method. For mobility scenarios, the overall identification accuracy is 96%. In addition, we perform security analysis of our proposed beam pattern fingerprinting system and PSD fingerprinting system by studying the feasibility of performing impersonation attacks. We design and implement an impersonation attack mechanism for mmWave wireless networks using state-of-the-art 60 GHz software defined radios. We discuss our findings and their implications on the security of the mmWave wireless networks.

cs.CR↗