A5039187782- Wireless Signal Modulation Classification
- Topic Modeling
- Advanced Graph Neural Networks
- Full-Duplex Wireless Communications
- Geophysical Methods and Applications
- Wireless Communication Security Techniques
- Terahertz technology and applications
- Digital Media Forensic Detection
- Radar Systems and Signal Processing
- Advanced Memory and Neural Computing
- Advanced Sensor and Control Systems
- Nuclear Physics and Applications
- Speech and Audio Processing
- Wireless Sensor Networks and IoT
- Rough Sets and Fuzzy Logic
- Advanced Image Processing Techniques
- Security in Wireless Sensor Networks
- Neural Networks and Reservoir Computing
- Image and Signal Denoising Methods
- UAV Applications and Optimization
- Bayesian Modeling and Causal Inference
- Industrial Technology and Control Systems
- Advanced SAR Imaging Techniques
- Epigenetics and DNA Methylation
- Complex Network Analysis Techniques
National University of Defense Technology
2020-2023
Beijing Information Science & Technology University
2021
PLA Electronic Engineering Institute
2017
Radio frequency fingerprints (RFFs) refer to the unique characteristics of signals transmitted by each emitter, which are valuable for physical layer security. Despite fact that deep learning based methods have shown significant advantages RFF extraction, they generally require large amount data avoid overfitting. To improve performance extraction with limited training samples, we propose an unsupervised pre-training method on masked autoencoding (MAE) learning. Specifically, neural network...
Specific Emitter Identification (SEI) is a key research problem in the field of information countermeasures. It one technologies required to be solved urgently target reconnaissance system. has ability distinguish between different individual radiation sources according varying characteristics emitter hardware within transmitted signals. In response lack scarcity among labeled samples specific identification, this paper proposes method combining multi-domain feature fusion and integrated...
The Internet of Things (IoT) is promising to transform a wide range fields. However, the open nature IoT makes it exposed cybersecurity threats, among which identity spoofing typical example. Physical layer authentication, identifies devices based on physical characteristics signals, serves as an effective way counteract spoofing. In this paper, we propose deep learning-based framework for open-set authentication devices. Specifically, additive angular margin softmax (AAMSoftmax) was...
In an electronic warfare-type scenario, optimal jamming strategy is vital important for a jammer who has restricted power and how to make the strategies quickly accurately put on agenda. this paper, we developed cognitive could learn with proposed algorithm-Greedy Bandits (GB). By interacting transmitter-receiver pairs continually, which also advantage of reinforcement learning theory, obtains physical layer parameters like signaling scheme, level on-off/pulsing. After constructing model,...
Specific emitter identification (SEI) is the process of identifying individual emitters by analyzing radio frequency emissions, based on fact that each device contains unique hardware imperfections.While majority previous research focuses obtaining features are discriminative, reliability rarely considered.For example, since characteristics same vary when it operating at different carrier frequencies, performance SEI approaches may degrade training data and test collected from with...
Specific emitter identification (SEI) refers to distinguishing emitters using individual features extracted from wireless signals. The current SEI methods have proven be accurate in tackling large labeled data sets at a high signal-to-noise ratio (SNR). However, their performance declines dramatically the presence of small samples and significant noise environment. To address this issue, we propose complex self-supervised learning scheme fully exploit unlabeled samples, comprised pretext...
Temporal knowledge graphs (KGs) have recently attracted increasing attention. The temporal KG forecasting task, which plays a crucial role in such applications as event prediction, predicts future links based on historical facts. However, current studies pay scant attention to the following two aspects. First, interpretability of models is manifested providing reasoning paths, an essential property path-based models. comparison paths these operated black-box fashion. Moreover, contemporary...
Knowledge graphs are structured representations of real world facts. However, they typically contain only a small subset all possible Link prediction is the task inferring missing facts based on existing ones. graph embedding, representing entities and relations in knowledge with high-dimensional vectors, has made significant progress link prediction. The tensor decomposition models an embedding family good performance previous do not consider problem attribute separation. These mainly...
Knowledge graphs (KGs) are collections of structured facts, which have recently attracted growing attention. Although there billions triples in KGs, they still incomplete. These incomplete knowledge bases will bring limitations to practical applications. Predicting new facts from the given is an increasingly important area. We investigate models based on logic rules this paper. This paper proposes HRER, a bottom-up rule learning for graph completion. First all, inspired by observation that...
Specific emitter identification (SEI) is extracting the features of received radio signals and determining individuals that generate signals. Although deep learning-based methods have been effectively applied for SEI, their performance declines dramatically with smaller number labeled training samples in presence significant noise. To address this issue, we propose an improved Bootstrap Your Own Late (BYOL) self-supervised learning scheme to fully exploit unlabeled samples, which comprises...
Knowledge graph embedding, representing entities and relations in the knowledge graphs with high-dimensional vectors, has made significant progress link prediction. More researchers have explored representational capabilities of models recent years. That is, they investigate better to fit symmetry/antisymmetry combination relationships. The current embedding are more inclined utilize identical vector for same entity various triples measure matching performance. observation that measuring...
Specific emitter identification (SEI) can distinguish single-radio transmitters using the subtle features of received waveform.Therefore, it is used extensively in both military and civilian fields.However, traditional method requires extensive prior knowledge time-consuming.Furthermore, imposes various effects associated with identifying communication radiation source signal complex environments.To solve problem weak robustness hand-crafted feature method, many scholars at home abroad have...
In order to solve the scarcity of labeled samples in actual combat scenarios makes it difficult converge deep learning model a complex electromagnetic environment. This paper combines residual network and complex-valued network, we proposes feature fusion that full use method has small amount data. First, need input reprocessed data into two networks. And then, regard features extracted by networks as real imaginary parts classifier train classifier. Finally, Using signal radiation source be...
Specific emitter identification (SEI) refers to the process of identifying individuals based on corresponding wireless signals. Although deep learning has been successfully applied in SEI, performance remains be improved when channel changes. In this paper, we suggest that a potential reason degradation is inadequacy model capacity. Therefore, Transformer, an advanced neural network architecture with large capacity, for channel-robust SEI. Experimental results show Transformer achieves...
Specific Emitter Identification (SEI) is the approach to identify emitter individuals using received wireless signals. Despite fact that deep learning has been successfully applied in SEI, performance still unsatisfying when receiver changes. In this paper, we introduce a domain adaptation method, namely Deep Adversarial Neural Network (DANN), for cross-receiver SEI. Furthermore, separated batch normalization (SepBN) proposed improve performance. Results of experiments real data show...
The convolutional neural network (CNN) shows excellent performances in various applications at the cost of huge storage and calculation overheads. To compress accelerate CNN, conventional methods usually prune unimportant filters under pre-selected standards, which should be artificially carefully designed according to specific application scenarios. In this paper, we propose a general CNN accelerating compressing method, i.e., parasitic mechanism (PAM)-based filter pruning algorithm. First,...
Abstract Emitter identification technology can distinguish the types of radiation sources and identify identity emitter. It has broad application prospects in both military civilian fields. The article mainly reviews source feature extraction methods for individual recent years, discusses advantages disadvantages manually extracted features based on deep learning. technical difficulties are summarized with respect to environment, number sources, performance algorithms, etc. Finally, points...
An online trajectory planner permits autonomous unmanned vehicles to maneuver in a changing mission scenario and varying environment. However, real-time plannings, task performance, completeness accuracy is seriously challenged by the latency during planning scheme computation. To address issue, we propose latency-correcting based on successive convex approximation (SCA), which aiming at offset planning. In this work, two types of are considered, i.e., prior known variable computation, where...