Xudong Zhong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Xudong Zhong has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
- Baoquan Ren
Systems Engineering Institute, AMS, PLA, Beijing, China
8 shared publications - Qijie Qian
Nanjing University of Posts and Telecommunications
6 shared publications - Zhen Qin
Department of Information and Communication, Noncommissioned Officer Academy of PAP, Hangzhou 310000, China
6 shared publications - Jiaqi Lin
Beijing University of Posts and Telecommunications
5 shared publications - Zibo Zhou
Systems Engineering Institute, AMS, PLA, Beijing, China
5 shared publications Baoquan Ren
4 shared publications
Latest Publications
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications
- Vol. 33 (5) , 27-35
Intelligent 6G Networks: From Connected Things to Connected Cognition—An Agentic-Native Framework
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications
- Vol. 33 (5) , 27-35
Intelligent 6G Networks: From Connected Things to Connected Cognition—An Agentic-Native Framework
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Magazine
- Vol. 9 (4) , 78-85
RIS-Empowered Satellite IoT: Bridging the Coverage-Efficiency Gap of Last-Mile Access and Sensing
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Magazine
- Vol. 9 (4) , 78-85
RIS-Empowered Satellite IoT: Bridging the Coverage-Efficiency Gap of Last-Mile Access and Sensing
- Open Access
- 5 May 2026
- Published by Elsevier in Digital Communications and Networks
Cognitive sovereignty in 6G: architecting a Neuro-Symbolic knowledge plane for verifiable agentic automation
- 5 May 2026
- Published by Elsevier in Digital Communications and Networks
Open AccessCognitive sovereignty in 6G: architecting a Neuro-Symbolic knowledge plane for verifiable agentic automation
- 23 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Green Communications and Networking
- Vol. 10, 2892-2906
Spatiotemporal Attention-Based Deep Reinforcement Learning for Joint Routing and Resource Management in Wireless Ad-Hoc Networks
Establishing high-quality communication routes in wireless ad-hoc networks remains challenging due to the strong coupling between routing strategies and resource management. To enable energy-efficient and robust networking, this paper proposes a novel cross-layer optimization framework based on spatiotemporal attention mechanism for joint routing and power management. The framework employs a deep reinforcement learning (DRL) method with hop-by-hop optimization to tackle the inherent partially observable Markov Decision Problem (POMDP) encountered during route establishment. At each node, a graph attention network (GAT) captures the irregular topology and dynamic interference relationships within its local neighborhood, while a Transformer encoder module models temporal evolution of channel and traffic conditions from historical observations. Priority experience replay is incorporated to improve sample efficiency and accelerate convergence. By effectively integrating spatial graph information and temporal context, the method significantly extends effective observability, thereby enabling superior interference mitigation and more efficient power management. Numerical simulations demonstrate that the proposed spatiotemporal attention-based approach outperforms conventional and DRL-based baselines in bottleneck throughput, transmission delay, and power efficiency. These results confirm the framework’s effectiveness, robustness and strong suitability for green communication in highly dynamic and resource-constrained wireless networks.
- 23 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Green Communications and Networking
- Vol. 10, 2892-2906
Spatiotemporal Attention-Based Deep Reinforcement Learning for Joint Routing and Resource Management in Wireless Ad-Hoc Networks
Establishing high-quality communication routes in wireless ad-hoc networks remains challenging due to the strong coupling between routing strategies and resource management. To enable energy-efficient and robust networking, this paper proposes a novel cross-layer optimization framework based on spatiotemporal attention mechanism for joint routing and power management. The framework employs a deep reinforcement learning (DRL) method with hop-by-hop optimization to tackle the inherent partially observable Markov Decision Problem (POMDP) encountered during route establishment. At each node, a graph attention network (GAT) captures the irregular topology and dynamic interference relationships within its local neighborhood, while a Transformer encoder module models temporal evolution of channel and traffic conditions from historical observations. Priority experience replay is incorporated to improve sample efficiency and accelerate convergence. By effectively integrating spatial graph information and temporal context, the method significantly extends effective observability, thereby enabling superior interference mitigation and more efficient power management. Numerical simulations demonstrate that the proposed spatiotemporal attention-based approach outperforms conventional and DRL-based baselines in bottleneck throughput, transmission delay, and power efficiency. These results confirm the framework’s effectiveness, robustness and strong suitability for green communication in highly dynamic and resource-constrained wireless networks.
- Open Access
- 1 April 2026
- Published by KEAI Publishing in Digital Communications and Networks
- Vol. 12 (4) , 618-629
Improved PHY-layer authentication utilizing multi-modal features for mmWave MIMO UAV-enabled systems
- 1 April 2026
- Published by KEAI Publishing in Digital Communications and Networks
- Vol. 12 (4) , 618-629
Open AccessImproved PHY-layer authentication utilizing multi-modal features for mmWave MIMO UAV-enabled systems
- Open Access
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Communications Society
- Vol. 7, 2430-2445
Attention-Enhanced Multi-Objective Deep Reinforcement Learning for Joint Power and Frequency Resource Allocation in Multi-Beam Satellite Communications
As a core component of Satellite-Terrestrial Integrated Networks (STINs), multi-beam satellite systems face prominent challenges in time-frequency resource allocation, including scarce spectrum resources, severe inter-beam interference, and conflicting multi-objective requirements. Deep reinforcement learning (DRL) provides an effective end-to-end decision-making framework with strong environmental perception capabilities. However, existing multi-objective DRL (MODRL) approaches for satellite resource allocation still exhibit shortcomings such as insufficient extraction of key state features, unbalanced multi-objective trade-offs, and weak anti-interference capabilities. To address these issues, this paper proposes an attention-enhanced multi-objective proximal policy optimization algorithm (Attention-MO-PPO). Firstly, a four-objective optimization model incorporating system throughput, user fairness, energy efficiency, and anti-interference performance is established, with improved channel state modeling via an interference-aware factor. Secondly, a dual attention mechanism integrating spatial and temporal attention is designed to enhance the capture of high-priority user channel features and dynamic interference patterns. Thirdly, an adaptive weight reward function based on Pareto optimality theory is developed to optimize the distribution uniformity of the Pareto solution set using the Shannon-Wiener diversity index. Finally, an adaptive adjustment mechanism for the clipping threshold is introduced to improve the algorithm’s training stability. Simulation results demonstrate that the proposed Attention-MO-PPO significantly outperforms existing MODRL and traditional methods in six key indicators: system throughput (increased by 23.7%), user fairness index (improved by 18.5%), energy efficiency (enhanced by 21.3%), interference suppression ratio (reduced by 41.8%), end-to-end delay (decreased by 27.6%), and packet loss rate (lowered by 55.0%). This paper provides a high-performance resource allocation solution for complex dynamic multi-beam satellite communication scenarios in STINs.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Communications Society
- Vol. 7, 2430-2445
Open AccessAttention-Enhanced Multi-Objective Deep Reinforcement Learning for Joint Power and Frequency Resource Allocation in Multi-Beam Satellite Communications
As a core component of Satellite-Terrestrial Integrated Networks (STINs), multi-beam satellite systems face prominent challenges in time-frequency resource allocation, including scarce spectrum resources, severe inter-beam interference, and conflicting multi-objective requirements. Deep reinforcement learning (DRL) provides an effective end-to-end decision-making framework with strong environmental perception capabilities. However, existing multi-objective DRL (MODRL) approaches for satellite resource allocation still exhibit shortcomings such as insufficient extraction of key state features, unbalanced multi-objective trade-offs, and weak anti-interference capabilities. To address these issues, this paper proposes an attention-enhanced multi-objective proximal policy optimization algorithm (Attention-MO-PPO). Firstly, a four-objective optimization model incorporating system throughput, user fairness, energy efficiency, and anti-interference performance is established, with improved channel state modeling via an interference-aware factor. Secondly, a dual attention mechanism integrating spatial and temporal attention is designed to enhance the capture of high-priority user channel features and dynamic interference patterns. Thirdly, an adaptive weight reward function based on Pareto optimality theory is developed to optimize the distribution uniformity of the Pareto solution set using the Shannon-Wiener diversity index. Finally, an adaptive adjustment mechanism for the clipping threshold is introduced to improve the algorithm’s training stability. Simulation results demonstrate that the proposed Attention-MO-PPO significantly outperforms existing MODRL and traditional methods in six key indicators: system throughput (increased by 23.7%), user fairness index (improved by 18.5%), energy efficiency (enhanced by 21.3%), interference suppression ratio (reduced by 41.8%), end-to-end delay (decreased by 27.6%), and packet loss rate (lowered by 55.0%). This paper provides a high-performance resource allocation solution for complex dynamic multi-beam satellite communication scenarios in STINs.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (20) , 1
Robust Multi-Domain Joint Optimization of Trajectory, Offloading, and Resources in UAV-MEC Networks
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (20) , 1
Robust Multi-Domain Joint Optimization of Trajectory, Offloading, and Resources in UAV-MEC Networks
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (17) , 1
Multi-Agent DRL with Dual-Stream Advantage Mixing for Anti-Jamming Resource Allocation
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (17) , 1
Multi-Agent DRL with Dual-Stream Advantage Mixing for Anti-Jamming Resource Allocation