Dongyang Xu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Dongyang Xu has a total of 74 co-authors. Here are the top co-authors by number of shared publications:
- Pinyi Ren
School of Information and Communications Engineering, Xi'an Jiaotong University, Xi'an 710049, China; Shaanxi Smart Networks and Ubiquitous Access Research Center, Xi'an 710049, China
17 shared publications - Tiantian Zhang
School of Cyber Science and Engineering, Xian Jiaotong University, Xian, China
15 shared publications - Keping Yu
Hosei University
11 shared publications - Mohsen Guizani
Mohamed bin Zayed University of Artificial Intelligence
8 shared publications James A. Ritcey
Department of Electrical and Computer Engineering, University of Washington, Seattle, WA, USA
7 shared publications- Lei Liu
Xidian University
6 shared publications
Latest Publications
- 6 October 2026
- Published by Springer Nature
- p. 48-59
Improved RT-DETR Model for Blind Parameter Estimation of FH Signals Under VLF Interference
- 6 October 2026
- Published by Springer Nature
- p. 48-59
Improved RT-DETR Model for Blind Parameter Estimation of FH Signals Under VLF Interference
- 21 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (10) , 16024-16038
Joint Trajectory and Beamforming Optimization for UAV-RIS-Empowered Multiuser Communication Networks: A Double Deep Q-Network Approach
Reconfigurable intelligent surface (RIS) assisted unmanned aerial vehicle (UAV) communication technology offers adaptable reflected beams and dynamic parameter adjustments, proving vital for managing the highly dynamic communication environments expected in wireless communications. However, a key challenge lies in addressing the dynamic and nonlinear problem of RIS phase shift design and UAV trajectory prediction. To ensure long-term fairness and enhance system performance, we propose an approach based on the double deep Q-network (DDQN), aiming to maximize the weighted sum rate across multiple users. Specifically, we employ an initial phase shift and predefined codebook to optimize the RIS phase shift at the current UAV position using DDQN. Further refinement is achieved through an extremum seeking algorithm. Then, under constraints encompassing RIS phase shift, allowable regions for UAV node, base station (BS) transmit power, and minimum user rate, we adopt DDQN to obtain predefined reward parameters jointly designed by fairness and throughput through interaction with dynamic and variable channel environment, and continuously update UAV position parameters to maximize rewards. The simulation results verify the efficacy of our joint optimization strategy, demonstrating optimal UAV location selection for UAV-RIS-empowered multiuser networks in a 2D plane.
- 21 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (10) , 16024-16038
Joint Trajectory and Beamforming Optimization for UAV-RIS-Empowered Multiuser Communication Networks: A Double Deep Q-Network Approach
Reconfigurable intelligent surface (RIS) assisted unmanned aerial vehicle (UAV) communication technology offers adaptable reflected beams and dynamic parameter adjustments, proving vital for managing the highly dynamic communication environments expected in wireless communications. However, a key challenge lies in addressing the dynamic and nonlinear problem of RIS phase shift design and UAV trajectory prediction. To ensure long-term fairness and enhance system performance, we propose an approach based on the double deep Q-network (DDQN), aiming to maximize the weighted sum rate across multiple users. Specifically, we employ an initial phase shift and predefined codebook to optimize the RIS phase shift at the current UAV position using DDQN. Further refinement is achieved through an extremum seeking algorithm. Then, under constraints encompassing RIS phase shift, allowable regions for UAV node, base station (BS) transmit power, and minimum user rate, we adopt DDQN to obtain predefined reward parameters jointly designed by fairness and throughput through interaction with dynamic and variable channel environment, and continuously update UAV position parameters to maximize rewards. The simulation results verify the efficacy of our joint optimization strategy, demonstrating optimal UAV location selection for UAV-RIS-empowered multiuser networks in a 2D plane.
- 13 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (10) , 15848-15865
Augmented V2V Radio Fingerprinting With Heterogeneous Federated Scattering Network for Autonomous Vehicle Road Cooperation
Augmented Internet of vehicle (IoV) network, providing natural infrastructure support for autonomous vehicle road cooperation (VRC), has been deemed as one of key enablers for autonomous VRC while simultaneously facing the challenge of impersonation, tampering or replay attack towards the high layer digital certificate for critical information. In this paper, we propose a heterogeneous scattering fingerprinting network (HSFNet) which amalgamates wavelet scattering and deep learning to achieve lightweight, rapid and reliable physical layer authentication by radio frequency fingerprinting (RFF) which employs the unique inherent hardware impairments within for vehicle-to everything (V2X) communication. Particularly, we first exploit a wavelet scattering network to extract RFF coefficients and eliminate redundancies within heterogeneous datasets. The establishment of a high-information-density datasets facilitates the efficient processing of extensive heterogeneous data, especially for resource-limited vehicles nodes. To improve the interpretability, we combine the multi-scale coefficients of input waveform and shapley values, formulating a complete interpretable framework and completing parameters tuning. Finally, considering heterogeneous applications in practical scenarios, we design a distributed federated learning framework which not only completes distributed training but also protects privacy with the resources-constrained intelligent IoV nodes. Experimental results demonstrate that the proposed scheme maintains excellent performance while utilizing only about 25.7% of original samples, significantly improving identification accuracy. Furthermore, the identification accuracy of federated learning ultimately converges to 98.1%.
- 13 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (10) , 15848-15865
Augmented V2V Radio Fingerprinting With Heterogeneous Federated Scattering Network for Autonomous Vehicle Road Cooperation
Augmented Internet of vehicle (IoV) network, providing natural infrastructure support for autonomous vehicle road cooperation (VRC), has been deemed as one of key enablers for autonomous VRC while simultaneously facing the challenge of impersonation, tampering or replay attack towards the high layer digital certificate for critical information. In this paper, we propose a heterogeneous scattering fingerprinting network (HSFNet) which amalgamates wavelet scattering and deep learning to achieve lightweight, rapid and reliable physical layer authentication by radio frequency fingerprinting (RFF) which employs the unique inherent hardware impairments within for vehicle-to everything (V2X) communication. Particularly, we first exploit a wavelet scattering network to extract RFF coefficients and eliminate redundancies within heterogeneous datasets. The establishment of a high-information-density datasets facilitates the efficient processing of extensive heterogeneous data, especially for resource-limited vehicles nodes. To improve the interpretability, we combine the multi-scale coefficients of input waveform and shapley values, formulating a complete interpretable framework and completing parameters tuning. Finally, considering heterogeneous applications in practical scenarios, we design a distributed federated learning framework which not only completes distributed training but also protects privacy with the resources-constrained intelligent IoV nodes. Experimental results demonstrate that the proposed scheme maintains excellent performance while utilizing only about 25.7% of original samples, significantly improving identification accuracy. Furthermore, the identification accuracy of federated learning ultimately converges to 98.1%.
- 2 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (1) , 24-38
Decentralized Federated Learning in LEO-Satellite-Based IoE Communications: Latency Optimization Under Reliability Constraints
Artificial intelligence-empowered low-Earth orbit (LEO) satellite networks have the great potential to provide robust and ubiquitous communications capabilities for Internet of Everything (IoE) services. Due to the intrinsic high decentralization, it is expected for LEO satellite nodes to deploy decentralized federated learning (DFL) for model training collaboratively while preserving data privacy. However, the challenges lie in the latency management of the DFL process under communications reliability constraints. To address the issue, we propose a novel DFL framework that incorporates halving and doubling to balance the load of networks. Considering the dependency of aggregation, the latency of DFL clients is recursively derived. Under constraints of energy consumption and network reliability, an optimization problem is formulated that aims to minimize the mean latency, solved using the particle swarm optimization algorithm. Additionally, we introduce a latency-aware scheduling strategy to further improve DFL latency efficiency by leveraging overlapping intersatellite links among DFL clients. Simulation results show that the proposed methods significantly accelerate the DFL process by up to 16.2% and enhance its efficiency compared to the alternatives.
- 2 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (1) , 24-38
Decentralized Federated Learning in LEO-Satellite-Based IoE Communications: Latency Optimization Under Reliability Constraints
Artificial intelligence-empowered low-Earth orbit (LEO) satellite networks have the great potential to provide robust and ubiquitous communications capabilities for Internet of Everything (IoE) services. Due to the intrinsic high decentralization, it is expected for LEO satellite nodes to deploy decentralized federated learning (DFL) for model training collaboratively while preserving data privacy. However, the challenges lie in the latency management of the DFL process under communications reliability constraints. To address the issue, we propose a novel DFL framework that incorporates halving and doubling to balance the load of networks. Considering the dependency of aggregation, the latency of DFL clients is recursively derived. Under constraints of energy consumption and network reliability, an optimization problem is formulated that aims to minimize the mean latency, solved using the particle swarm optimization algorithm. Additionally, we introduce a latency-aware scheduling strategy to further improve DFL latency efficiency by leveraging overlapping intersatellite links among DFL clients. Simulation results show that the proposed methods significantly accelerate the DFL process by up to 16.2% and enhance its efficiency compared to the alternatives.
- 19 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- Vol. 39 (5) , 145-154
Digital Twin Enabled 6G Cell-Free Multi-Function Reconfigurable Metasurfaces Communications: Concepts, Challenges, and Future Directions
Cell-free massive multiple-input multiple-output (CF-mMIMO) and multi-function reconfigurable metasurfaces are two potential technologies for coping with 6G demands. Their coexistence, also called cell-free reconfigurable metasurfaces (CFRM) communications, contribute to enhancing wireless network performances in an energy-and-hardware-efficient manner but shows great voracious appetite for data, optimization and resources. Digital twin (DT) provides a hotbed for the development of CFRM communications. This article aims to present how DT and CFRM communications are organically merged as an integrated whole, which enable 6G deployment from a perspective of three typical cases, including intelligent reconfigurable holographic surface (RHS)/reconfigurable intelligent surface (RIS) deployment optimization, hybrid beamforming optimization and intelligent model/resource orchestration. A real-world DT enabled CFRM communications application is presented, and simulation results have demonstrated that DT enabled CFRM can provide remarkable evaluate the implementation platform for subarray reconfigurable metasurfaces based on reinforcement learning to improve the energy efficiency of CFRM. Finally, we conclude the article with potential challenges of RHS/RISs and several future research topics.
- 19 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- Vol. 39 (5) , 145-154
Digital Twin Enabled 6G Cell-Free Multi-Function Reconfigurable Metasurfaces Communications: Concepts, Challenges, and Future Directions
Cell-free massive multiple-input multiple-output (CF-mMIMO) and multi-function reconfigurable metasurfaces are two potential technologies for coping with 6G demands. Their coexistence, also called cell-free reconfigurable metasurfaces (CFRM) communications, contribute to enhancing wireless network performances in an energy-and-hardware-efficient manner but shows great voracious appetite for data, optimization and resources. Digital twin (DT) provides a hotbed for the development of CFRM communications. This article aims to present how DT and CFRM communications are organically merged as an integrated whole, which enable 6G deployment from a perspective of three typical cases, including intelligent reconfigurable holographic surface (RHS)/reconfigurable intelligent surface (RIS) deployment optimization, hybrid beamforming optimization and intelligent model/resource orchestration. A real-world DT enabled CFRM communications application is presented, and simulation results have demonstrated that DT enabled CFRM can provide remarkable evaluate the implementation platform for subarray reconfigurable metasurfaces based on reinforcement learning to improve the energy efficiency of CFRM. Finally, we conclude the article with potential challenges of RHS/RISs and several future research topics.
- 18 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (5) , 4935-4952
Deep Federated Fractional Scattering Network for Heterogeneous Edge Internet of Vehicles Fingerprinting: Theory and Implementation
With the rapid development of distributed edge intelligence (DEI) within Internet of vehicle (IoV) network, it is required to support heterogeneous rapid, reliable and lightweight authentication which prevents eavesdropping, tampering and replay attacks. Radio Frequency Fingerprinting (RFF), which leverages unique and tamper-proof hardware characteristics, is an emerging deep learning based physical layer technology poised to achieve excellent authentication within DEI enhanced heterogeneous IoV. However, centralized collection of critical datasets will bring severe privacy concerns as well as huge communication overheads towards resources-constrained IoV nodes. In this paper, we propose a deep federated fractional scattering fingerprinting network (FFSFNet) which amalgamates fractional wavelet scattering and federated learning to achieve excellent identification. Particularly, we first exploit fractional wavelet scattering to extract RFF characteristics from non-stationary waveform, eliminate redundancies and enhance interpretability. To improve the training efficiency and privacy protection capability, we design a novel federated framework, which not only completes distributed training, reduces overhead but also protects privacy. Furthermore, we conducted a comprehensive comparative analysis of different model quantization schemes and validated the proposed scheme with field programmable gate array (FPGA) accelerators. Experimental results demonstrate that the proposed FFSFNet can maintain excellent identification performance with only 5.08% of original samples. The model size and inference latency can be effectively improved by quantization with limited degradation. Moreover, the identification testing accuracy of FFSFNet can eventually converge to 99.4% with 0.64ms inference latency per sample.
- 18 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (5) , 4935-4952
Deep Federated Fractional Scattering Network for Heterogeneous Edge Internet of Vehicles Fingerprinting: Theory and Implementation
With the rapid development of distributed edge intelligence (DEI) within Internet of vehicle (IoV) network, it is required to support heterogeneous rapid, reliable and lightweight authentication which prevents eavesdropping, tampering and replay attacks. Radio Frequency Fingerprinting (RFF), which leverages unique and tamper-proof hardware characteristics, is an emerging deep learning based physical layer technology poised to achieve excellent authentication within DEI enhanced heterogeneous IoV. However, centralized collection of critical datasets will bring severe privacy concerns as well as huge communication overheads towards resources-constrained IoV nodes. In this paper, we propose a deep federated fractional scattering fingerprinting network (FFSFNet) which amalgamates fractional wavelet scattering and federated learning to achieve excellent identification. Particularly, we first exploit fractional wavelet scattering to extract RFF characteristics from non-stationary waveform, eliminate redundancies and enhance interpretability. To improve the training efficiency and privacy protection capability, we design a novel federated framework, which not only completes distributed training, reduces overhead but also protects privacy. Furthermore, we conducted a comprehensive comparative analysis of different model quantization schemes and validated the proposed scheme with field programmable gate array (FPGA) accelerators. Experimental results demonstrate that the proposed FFSFNet can maintain excellent identification performance with only 5.08% of original samples. The model size and inference latency can be effectively improved by quantization with limited degradation. Moreover, the identification testing accuracy of FFSFNet can eventually converge to 99.4% with 0.64ms inference latency per sample.
- 16 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 11 (6) , 6259-6271
Joint Data Allocation and LSTM-Based Server Selection With Parallelized Federated Learning in LEO Satellite IoT Networks
Low earth orbit (LEO) satellite networks have emerged as a promising field for distributed Internet of Things (IoT) devices, particularly in latency-tolerant applications. Federated learning (FL) is implemented in LEO satellite IoT networks to preserve data privacy and facilitate machine learning (ML). However, the user who spends the longest time significantly hampers FL efficiency and degrades the Quality-of-Service (QoS), potentially leading to irreparable damage. To address this challenge, we propose a joint data allocation and server selection strategy based on long short-term memory (LSTM) with parallelized FL in LEO satellite IoT networks. Herein, data-parallel learning is utilized, allowing multiple users to collaboratively train ML networks to minimize latency. Moreover, server selection takes into account signal propagation delays as well as traffic loads forecasted by an LSTM network, thereby improving the efficiency even further. Specifically, the strategies are formulated as optimization problems and tackled using a line search sequential quadratic programming (SQP) method and a multiple-objective particle swarm optimization (MOPSO) algorithm. Simulation results show the effectiveness of the proposed strategy in reducing total latency and enhancing the efficiency of FL in LEO satellite IoT networks compared to the alternatives.
- 16 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 11 (6) , 6259-6271
Joint Data Allocation and LSTM-Based Server Selection With Parallelized Federated Learning in LEO Satellite IoT Networks
Low earth orbit (LEO) satellite networks have emerged as a promising field for distributed Internet of Things (IoT) devices, particularly in latency-tolerant applications. Federated learning (FL) is implemented in LEO satellite IoT networks to preserve data privacy and facilitate machine learning (ML). However, the user who spends the longest time significantly hampers FL efficiency and degrades the Quality-of-Service (QoS), potentially leading to irreparable damage. To address this challenge, we propose a joint data allocation and server selection strategy based on long short-term memory (LSTM) with parallelized FL in LEO satellite IoT networks. Herein, data-parallel learning is utilized, allowing multiple users to collaboratively train ML networks to minimize latency. Moreover, server selection takes into account signal propagation delays as well as traffic loads forecasted by an LSTM network, thereby improving the efficiency even further. Specifically, the strategies are formulated as optimization problems and tackled using a line search sequential quadratic programming (SQP) method and a multiple-objective particle swarm optimization (MOPSO) algorithm. Simulation results show the effectiveness of the proposed strategy in reducing total latency and enhancing the efficiency of FL in LEO satellite IoT networks compared to the alternatives.
- 9 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (2) , 3474-3479
Cooperative IRS and IOS-Aided Finite Blocklength Covert Communications in NOMA Networks
This paper proposes a finite blocklength covert non-orthogonal multiple access (NOMA) communication scheme based on cooperative intelligent reflecting surfaces (IRS) and intelligent omni-surfaces (IOS). To achieve covert transmission in the NOMA network, in the proposed scheme, the IRS switches between the IRS and IOS mechanisms to create uncertainty. To analyze the performance of this scheme, we first characterize the statistical properties of the users' cascaded channels. Then, we derive the closed-form expressions of the expectation of the Kullback-Leibler (KL) divergence and the average block error rates (BLERs) of NOMA users. To balance covertness and reliability, we formulate an optimization problem to maximize the effective covert throughput (ECT) subject to the reliability and delay constraints and the level of covertness. Via this optimization, we then obtain closed-form expressions for optimal power and blocklength. Simulation results demonstrate that the ECT of the proposed scheme is much higher than that of benchmarks.
- 9 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (2) , 3474-3479
Cooperative IRS and IOS-Aided Finite Blocklength Covert Communications in NOMA Networks
This paper proposes a finite blocklength covert non-orthogonal multiple access (NOMA) communication scheme based on cooperative intelligent reflecting surfaces (IRS) and intelligent omni-surfaces (IOS). To achieve covert transmission in the NOMA network, in the proposed scheme, the IRS switches between the IRS and IOS mechanisms to create uncertainty. To analyze the performance of this scheme, we first characterize the statistical properties of the users' cascaded channels. Then, we derive the closed-form expressions of the expectation of the Kullback-Leibler (KL) divergence and the average block error rates (BLERs) of NOMA users. To balance covertness and reliability, we formulate an optimization problem to maximize the effective covert throughput (ECT) subject to the reliability and delay constraints and the level of covertness. Via this optimization, we then obtain closed-form expressions for optimal power and blocklength. Simulation results demonstrate that the ECT of the proposed scheme is much higher than that of benchmarks.