A5044419947- Speech and Audio Processing
- Parallel Computing and Optimization Techniques
- Interconnection Networks and Systems
- Advanced Wireless Communication Techniques
- Underwater Vehicles and Communication Systems
- Advanced MIMO Systems Optimization
- Music and Audio Processing
- Indoor and Outdoor Localization Technologies
- Animal Vocal Communication and Behavior
- Embedded Systems Design Techniques
- Advanced Fiber Optic Sensors
- Advanced Wireless Communication Technologies
- Vehicular Ad Hoc Networks (VANETs)
- Error Correcting Code Techniques
- High-Velocity Impact and Material Behavior
- Advanced Fiber Laser Technologies
- Advanced Memory and Neural Computing
- Neural Networks and Applications
- Surface Treatment and Residual Stress
- Video Surveillance and Tracking Methods
- Advanced Neural Network Applications
- Power Line Communications and Noise
- Autonomous Vehicle Technology and Safety
- Sparse and Compressive Sensing Techniques
- Vehicle License Plate Recognition
Shanghai University
2017-2025
Xiamen University
2023-2024
University of Lincoln
2023
Aviation Industry Corporation of China (China)
2023
University of California, Berkeley
2018-2022
City University of Hong Kong
2018-2021
ORCID
2021
University of West Florida
2019-2020
Shenyang Institute of Automation
2020
Chinese Academy of Sciences
2020
Recently, nonuniform linear arrays (e.g., coprime/nested array) have attracted great attention of researchers in array signal processing field due to its ability generate virtual difference coarrays. In the design, a critical problem is where place sensors for optimal performance aiming maximum degree freedom capacity and minimum mutual coupling ratio, simultaneously. An augmented nested concept proposed by splitting dense subarray into several parts, which can be rearranged at two sides...
We demonstrate a novel type of distributed optical fiber acoustic sensor, with the ability to detect and retrieve actual temporal waveforms multiple vibration events that occur simultaneously at different positions along fiber. The system is realized via dual-pulse phase-sensitive time-domain reflectometry, waveform retrieved by heterodyne phase demodulation. Experimental results show has background noise level as low 8.91×10-4 rad/√Hz demodulation signal-to-noise ratio 49.17 dB 1 kHz, can...
Orthogonal frequency division multiplexing (OFDM) provides a promising modulation technique for underwater acoustic (UWA) communication systems. It is indispensable to obtain channel state information estimation handle the various distortions and interferences. However, conventional methods such as least square (LS), minimum mean error (MMSE) back propagation neural network (BPNN) cannot be directly applied UWA-OFDM systems, since complicated multipath channels may cause serious decline in...
Environmental sound classification (ESC) is a challenging problem due to the complexity of sounds. The performance heavily dependent on effectiveness representative features extracted from environmental However, ESC often suffers semantically irrelevant frames and silent frames. In order deal with this, we employ frame-level attention model focus relevant salient Specifically, first propose convolutional recurrent neural network learn spectro-temporal temporal correlations. Then, extend our...
Environmental sound classification (ESC) is a challenging problem due to the complex temporal structure and diverse energy modulation patterns of environmental sounds. In order deal with former, attention mechanism originally adopted focus on informative frames. However, no existing works pay latter problem. this paper, we consider role convolution filters in detecting propose channel semantically relevant channels generated by corresponding filters. Furthermore, incorporate enhance...
The development of flexible lithium-ion batteries (LIBs) imposes demands on energy density and high mechanical durability simultaneously. Due to the limited deformability electrodes, as well flat smooth surface metal current collectors, stable/durable/reliable contact between electrode materials collectors remains a challenge, in particular, for electrodes with loading mass heavily deformed batteries. Binders play an essential role binding particles adhering them collectors. Herein, inspired...
Precise indoor localization is an increasingly demanding requirement for various emerging applications, like Virtual/Augmented reality and personalized advertising. Current environments are equipped with pluralities of WiFi access points (APs), whose deployment expected to be massive in the future enabling highly precise approaches. Though conventional model-based schemes have achieved sub-meter level accuracy by fusing multiple channel state information (CSI) observations, corresponding...
Common-mode noises (CMNs) are frequently encountered in all types of fiber sensor systems. Through involving a reference interferometer, here we report the application differential delay heterodyne scheme to suppress CMNs fiber-optic interferometric The contributions shared by both sensing and interferometers further suppressed demodulation. For first time, develop full theoretical model describing this method, experimental results also presented confirm theory. Theoretical analysis shows...
Heterogeneous network (HetNet) has been proposed as a promising solution for handling the wireless traffic explosion in future fifth-generation (5G) system. In this paper, joint subchannel and power allocation problem is formulated HetNets to maximize energy efficiency (EE). By decomposing original into classification subproblem regression subproblem, convolutional neural (CNN) based approach developed obtain decisions on with much lower complexity than conventional iterative methods....
Wireless localization for mobile device has attracted more and interests by increasing the demand location based services. Fingerprint-based is promising, especially in non-Line-of-Sight (NLoS) or rich scattering environments, such as urban areas indoor scenarios. In this paper, we propose a novel fingerprint-based technique on deep learning framework under commercial long term evolution (LTE) systems. Specifically, develop software defined user equipment to collect real time channel state...
In the 5G communication systems, a hybrid approach to support polar codes for control plane and LDPC data has been identified as channel coding solution enhanced mobile broadband (eMBB) scenario. One of major challenges implement this is design powerful decoders at terminal side. Inspired from useful machine learning based decoder, we proposed deep unified polar-LDPC by concatenating an indicator section. Through numerical experiments, show that neural network (DNN) decoding architecture can...
Multiple-input multiple-output (MIMO) system is the key technology for long term evolution (LTE) and 5G. The information detection problem at receiver side in general difficult due to imbalance of decoding complexity accuracy within conventional methods. Hence, a deep learning based efficient MIMO approach proposed this paper. In our work, we use neural network directly get mapping function received signals, channel matrix transmitted bit streams. Then, compare end-to-end using with methods...
Pollution management has received significant critical attention in the leather industry, and effective solutions involve minimizing pollutant discharge during production optimizing waste utilization through reuse recycling. This study focused on developing an enzyme unhairing system to reduce sulfide pollution production, which utilized recombinant KerT protease with minimal collagenolytic activity avoid over-degrading hide collagen fibers. The properties of were systematically analyzed...
Purpose The modeling of interphase forces plays a significant role in the numerical simulation gas–liquid flow rotodynamic multiphase pump, which deserves detailed study. Design/methodology/approach Numerical analysis is conducted to estimate influence forces, including drag force, lift virtual mass wall lubrication force and turbulent dispersion force. Findings results show that magnitude can be sorted by: > relations between velocity difference also gas volume fraction are revealed....
The influence factors of underwater acoustic propagation are the foundation communication. In this paper, a realistic channel is modeled in South China Sea. Three kinds typical sea surface models with curvilinear boundaries selected for investigation by BELLHOP ray model. Under different restrictions, characteristics model analyzed contrastively changing types, sending and receiving depths seasonal sound speed profiles. experimental results present that (1) types surfaces have little effect...