Xiongyan Tang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiongyan Tang has a total of 66 co-authors. Here are the top co-authors by number of shared publications:
- Lexi Xu
China United Network Communications Corporation
6 shared publications - Yi Yue
China Unicom Research Institute, Beijing, China
5 shared publications Chang Cao
China Unicom Research Institute
3 shared publications- Chang Cao
China Unicom, Beijing, China
3 shared publications Min Zhang
State Key Laboratory of Information Photonics and Optical Communications, Beijing University of Posts and Telecommunications, Beijing, China
3 shared publications- Wencong Yang
China Unicom Research Institute, BeiJing, China and National Engineering Research Center of Next Generation Internet Broadband Service Application, Beijing, China
3 shared publications
Latest Publications
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (11) , 19659-19674
AvailGuard: An Availability-Aware VNF Deployment Framework Based on Traffic Forecasting in SDN/NFV-Enabled Networks
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (11) , 19659-19674
AvailGuard: An Availability-Aware VNF Deployment Framework Based on Traffic Forecasting in SDN/NFV-Enabled Networks
- 15 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. PP (07338724) , 1-16
Implementation of Digital Twin for Optical Mesh Networks: Field Demonstration and Case Studies
The integration of digital twin (DT) technology in optical networks has profoundly transformed various operational aspects, including design, deployment, prediction, optimization, and network upgrades. While most DT verifications have been limited to controlled laboratory environments, there is a pressing need to extend this technology to field-deployed optical networks. This requirement is especially crucial for supporting large-scale mesh networks and high-capacity transmission like 400 or 800 Gbps, which operate in non-ideal conditions and present significant challenges for accurate DT implementation. We proposed a practically executable four-step approach for constructing DT models of field-deployed optical mesh networks, involving physics-guided sequential parameter refinement to ensure precision and adaptability to real-time conditions. Validations of our approach through field-trial testbed demonstrated its accuracy, effectiveness, adaptability, and robustness, exemplified by five scenarios. Our results showed that the DT model could precisely predict network performance, with absolute errors in signal-to-noise ratio (SNR) of less than 0.57 dB, and can continuously adapt to upgrade cases such as 800 Gbps transmission. The methodology bridges theoretical advancements with practical applications, offering enhanced efficiency and resilience in real-world optical network deployments.
- 15 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. PP (07338724) , 1-16
Implementation of Digital Twin for Optical Mesh Networks: Field Demonstration and Case Studies
The integration of digital twin (DT) technology in optical networks has profoundly transformed various operational aspects, including design, deployment, prediction, optimization, and network upgrades. While most DT verifications have been limited to controlled laboratory environments, there is a pressing need to extend this technology to field-deployed optical networks. This requirement is especially crucial for supporting large-scale mesh networks and high-capacity transmission like 400 or 800 Gbps, which operate in non-ideal conditions and present significant challenges for accurate DT implementation. We proposed a practically executable four-step approach for constructing DT models of field-deployed optical mesh networks, involving physics-guided sequential parameter refinement to ensure precision and adaptability to real-time conditions. Validations of our approach through field-trial testbed demonstrated its accuracy, effectiveness, adaptability, and robustness, exemplified by five scenarios. Our results showed that the DT model could precisely predict network performance, with absolute errors in signal-to-noise ratio (SNR) of less than 0.57 dB, and can continuously adapt to upgrade cases such as 800 Gbps transmission. The methodology bridges theoretical advancements with practical applications, offering enhanced efficiency and resilience in real-world optical network deployments.
- 13 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 3625-3640
A Two-Timescale Resource Allocation Method Based on Deep Reinforcement Learning for 6G Networks
With the rapid development of artificial intelligence and the dramatic growth of communication services, the sixth-generation (6G) wireless network needs to handle communication tasks more flexibly and efficiently, significantly exacerbating the challenge of resource allocation. For the access network scenarios in 6G networks, the existing single-layer reinforcement learning resource allocation algorithms are hard to satisfy the diverse demands of users due to the complex and variable state space. Therefore, we propose a reinforcement learning-based two-timescale resource allocation scheme, aiming to jointly enhance the quality of service and system resource utilization. The proposed method comprises an upper-layer controller that allocates network resources to lower-layer controllers on a large time scale. Then, lower-layer controllers refine the resources based on user service types on a smaller time scale. To implement the proposed two-timescale allocation scheme, we propose a two-layer reinforcement learning framework consisting of a deep deterministic policy gradient (DDPG) and a dueling deep Q network (Dueling-DQN). Furthermore, recognizing that coupling multiple reinforcement learning processes may slow down algorithm convergence, we employ asynchronous training, transfer learning, and prediction-based action space simplification to expedite the model’s convergence speed. Finally, we build a prototyping network to verify the performance of the proposed small-timescale and the large-timescale allocation algorithms. Our proposed scheme demonstrates significant improvements in both resource utilization and quality of service compared to existing schemes.
- 13 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 3625-3640
A Two-Timescale Resource Allocation Method Based on Deep Reinforcement Learning for 6G Networks
With the rapid development of artificial intelligence and the dramatic growth of communication services, the sixth-generation (6G) wireless network needs to handle communication tasks more flexibly and efficiently, significantly exacerbating the challenge of resource allocation. For the access network scenarios in 6G networks, the existing single-layer reinforcement learning resource allocation algorithms are hard to satisfy the diverse demands of users due to the complex and variable state space. Therefore, we propose a reinforcement learning-based two-timescale resource allocation scheme, aiming to jointly enhance the quality of service and system resource utilization. The proposed method comprises an upper-layer controller that allocates network resources to lower-layer controllers on a large time scale. Then, lower-layer controllers refine the resources based on user service types on a smaller time scale. To implement the proposed two-timescale allocation scheme, we propose a two-layer reinforcement learning framework consisting of a deep deterministic policy gradient (DDPG) and a dueling deep Q network (Dueling-DQN). Furthermore, recognizing that coupling multiple reinforcement learning processes may slow down algorithm convergence, we employ asynchronous training, transfer learning, and prediction-based action space simplification to expedite the model’s convergence speed. Finally, we build a prototyping network to verify the performance of the proposed small-timescale and the large-timescale allocation algorithms. Our proposed scheme demonstrates significant improvements in both resource utilization and quality of service compared to existing schemes.
- Open Access
- 8 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 63-68
Adaptive Service Function Chaining with Generative Intelligence in Polymorphic Networks
- 8 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 63-68
Open AccessAdaptive Service Function Chaining with Generative Intelligence in Polymorphic Networks
- 30 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (4) , 5050-5064
Fairness-Aware Overtaking Decision Optimization for Mixed Connected and Connectionless Vehicles
In intelligent transportation systems, the ability to make precise and efficient lane-changing overtaking decisions is essential for improving traffic flow, safety, and overall efficiency. However, the coexistence of both connected and non-connected vehicles, driven by the high cost of full deployment and the incomplete global adoption of standardized communication standards, has led to the emergence of Mixed Connected and Connectionless Vehicles (MCCV) scenarios. These scenarios complicate lane-changing overtaking decisions, as the unpredictability of connectionless vehicles, particularly in dense traffic, poses significant challenges and increases safety risks. Furthermore, incorporating fairness into decision-making is vital to ensure equitable treatment of all vehicles, which is key to improving road safety and traffic efficiency. To address these challenges, we propose a fairness-aware overtaking decision optimization method for MCCV scenarios, which aims to enhance fairness while improving safety and efficiency in vehicle decision-making. First, a Bayesian network-based fairness assessment method is introduced to quantify fairness under limited data conditions by modeling the probabilistic relationships between vehicle behaviors and fairness outcomes. Second, we develop a left-lane availability detection mechanism based on adaptive maneuver tree search and a fast-lane speed recommendation mechanism grounded in traffic flow analysis. These mechanisms enhance compliance with traffic regulations and improve efficiency by dynamically assessing lane conditions and providing a real-time speed recommendation based on traffic density and flow. Finally, we incorporate a K-Nearest Neighbor (KNN)-enhanced deep reinforcement learning approach, which integrates a parameterized dueling deep recurrent Q-network with KNN-enhanced experience replay. This approach effectively copes with rare but critical driving conditions, such as unpredictable vehicle behaviors or sudden traffic dynamics, improving decision-making reliability in various traffic scenarios. Extensive simulations demonstrate that the proposed method significantly enhances the fairness, efficiency, and safety of lane-changing overtaking decisions in MCCV scenarios, effectively addressing the unpredictability of mixed-vehicle interactions.
- 30 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (4) , 5050-5064
Fairness-Aware Overtaking Decision Optimization for Mixed Connected and Connectionless Vehicles
In intelligent transportation systems, the ability to make precise and efficient lane-changing overtaking decisions is essential for improving traffic flow, safety, and overall efficiency. However, the coexistence of both connected and non-connected vehicles, driven by the high cost of full deployment and the incomplete global adoption of standardized communication standards, has led to the emergence of Mixed Connected and Connectionless Vehicles (MCCV) scenarios. These scenarios complicate lane-changing overtaking decisions, as the unpredictability of connectionless vehicles, particularly in dense traffic, poses significant challenges and increases safety risks. Furthermore, incorporating fairness into decision-making is vital to ensure equitable treatment of all vehicles, which is key to improving road safety and traffic efficiency. To address these challenges, we propose a fairness-aware overtaking decision optimization method for MCCV scenarios, which aims to enhance fairness while improving safety and efficiency in vehicle decision-making. First, a Bayesian network-based fairness assessment method is introduced to quantify fairness under limited data conditions by modeling the probabilistic relationships between vehicle behaviors and fairness outcomes. Second, we develop a left-lane availability detection mechanism based on adaptive maneuver tree search and a fast-lane speed recommendation mechanism grounded in traffic flow analysis. These mechanisms enhance compliance with traffic regulations and improve efficiency by dynamically assessing lane conditions and providing a real-time speed recommendation based on traffic density and flow. Finally, we incorporate a K-Nearest Neighbor (KNN)-enhanced deep reinforcement learning approach, which integrates a parameterized dueling deep recurrent Q-network with KNN-enhanced experience replay. This approach effectively copes with rare but critical driving conditions, such as unpredictable vehicle behaviors or sudden traffic dynamics, improving decision-making reliability in various traffic scenarios. Extensive simulations demonstrate that the proposed method significantly enhances the fairness, efficiency, and safety of lane-changing overtaking decisions in MCCV scenarios, effectively addressing the unpredictability of mixed-vehicle interactions.
- 2 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Technology Letters
- Vol. 37 (23) , 1381-1384
Real-Time Single-Carrier 800 Gbit/s S+C+L Band Optical Transmission Over 300 km Single-Mode Fiber
In this letter, we demonstrate a real-time S+C+L band optical transmission system with a total transmission capacity of 84 Tbit/s. 105 wavelength division multiplexing (WDM) channels with 150 GHz grid are used to transmit 135 GBaud probabilistic constellation shaping 16 quadrature amplitude modulation (PCS-16QAM) signals. The system only uses doped fiber amplifiers for signal amplification. We adopt self-developed S-band integrated tunable laser assembly (ITLA) with high output power to generate S-band signal. The real-time net data rate of S+C+L band is 800 Gbit/s and the total transmission distance is 300 km. To the best of our knowledge, it is currently the highest real-time net data rate reported in the S band and the farthest transmission distance of real-time S+C+L band transmission systems.
- 2 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Technology Letters
- Vol. 37 (23) , 1381-1384
Real-Time Single-Carrier 800 Gbit/s S+C+L Band Optical Transmission Over 300 km Single-Mode Fiber
In this letter, we demonstrate a real-time S+C+L band optical transmission system with a total transmission capacity of 84 Tbit/s. 105 wavelength division multiplexing (WDM) channels with 150 GHz grid are used to transmit 135 GBaud probabilistic constellation shaping 16 quadrature amplitude modulation (PCS-16QAM) signals. The system only uses doped fiber amplifiers for signal amplification. We adopt self-developed S-band integrated tunable laser assembly (ITLA) with high output power to generate S-band signal. The real-time net data rate of S+C+L band is 800 Gbit/s and the total transmission distance is 300 km. To the best of our knowledge, it is currently the highest real-time net data rate reported in the S band and the farthest transmission distance of real-time S+C+L band transmission systems.
- Open Access
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 249-251
Efficient Large-scale Model Training with Disaggregated Storage and Computing Architecture
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 249-251
Open AccessEfficient Large-scale Model Training with Disaggregated Storage and Computing Architecture
- Open Access
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 277-278
GenAI-SFC: A GenAI-assisted Approach for SFC Provision in 6G Intelligent Networks
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 277-278
Open AccessGenAI-SFC: A GenAI-assisted Approach for SFC Provision in 6G Intelligent Networks
- 2 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (15) , 32132-32147
Joint Beam Hopping and Precoding for Dense LEO Satellite Communication Systems
Beam hopping (BH) is a widely adopted technique in multibeam satellite communication systems, and it can effectively improve the system capacity. However, conventional BH with full frequency reuse requires spatial isolation to avoid interbeam interference, and it will impose restrictions on the flexibility of BH and achievable system capacity. By combining the BH and multibeam precoding, it is beneficial to satisfy the uneven traffic demands and keep the flexibility of beam management. Different from the existing works that focus on the joint BH and precoding for a single satellite, we investigate the joint multisatellite cell association (MCA), BH pattern design and multibeam precoding problem for dense low earth orbit (LEO) satellite communication systems. To tackle the complex problem, we decompose it into two subproblems. First, a many-to-one matching-based MCA algorithm is proposed, which balances the load among satellites and mitigates interference among activated cells. Second, many-to-many matching-based BH pattern design and quadratic transform-based precoding algorithms are proposed and alternately optimized to solve the joint BH and precoding problem. Lastly, a multisatellite joint BH and precoding algorithm based on alternating optimization (MJBHPAO) is proposed to solve the entire problem. The effectiveness of the proposed algorithm is verified with various parameters, and simulation results show that the proposed algorithm performs better than the referred schemes.
- 2 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (15) , 32132-32147
Joint Beam Hopping and Precoding for Dense LEO Satellite Communication Systems
Beam hopping (BH) is a widely adopted technique in multibeam satellite communication systems, and it can effectively improve the system capacity. However, conventional BH with full frequency reuse requires spatial isolation to avoid interbeam interference, and it will impose restrictions on the flexibility of BH and achievable system capacity. By combining the BH and multibeam precoding, it is beneficial to satisfy the uneven traffic demands and keep the flexibility of beam management. Different from the existing works that focus on the joint BH and precoding for a single satellite, we investigate the joint multisatellite cell association (MCA), BH pattern design and multibeam precoding problem for dense low earth orbit (LEO) satellite communication systems. To tackle the complex problem, we decompose it into two subproblems. First, a many-to-one matching-based MCA algorithm is proposed, which balances the load among satellites and mitigates interference among activated cells. Second, many-to-many matching-based BH pattern design and quadratic transform-based precoding algorithms are proposed and alternately optimized to solve the joint BH and precoding problem. Lastly, a multisatellite joint BH and precoding algorithm based on alternating optimization (MJBHPAO) is proposed to solve the entire problem. The effectiveness of the proposed algorithm is verified with various parameters, and simulation results show that the proposed algorithm performs better than the referred schemes.
- 24 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 24 (8) , 6945-6960
ALANINE: A Novel Decentralized Personalized Federated Learning for Heterogeneous LEO Satellite Constellation
Low Earth Orbit (LEO) satellite constellations have seen significant growth and functional enhancement in recent years, which integrates various capabilities like communication, navigation, and remote sensing. However, the heterogeneity of data collected by different satellites and the problems of efficient inter-satellite collaborative computation pose significant obstacles to realizing the potential of these constellations. Existing approaches struggle with data heterogeneity, varing image resolutions, and the need for efficient on-orbit model training. To address these challenges, we propose a novel decentralized PFL framework, namely, A Novel DecentraLized PersonAlized Federated Learning for HeterogeNeous LEO SatellIte CoNstEllation (ALANINE). ALANINE incorporates decentralized FL (DFL) for satellite image Super Resolution (SR), which enhances input data quality. Then it utilizes PFL to implement a personalized approach that accounts for unique characteristics of satellite data. In addition, the framework employs advanced model pruning to optimize model complexity and transmission efficiency. The framework enables efficient data acquisition and processing while improving the accuracy of PFL image processing models. Simulation results demonstrate that ALANINE exhibits superior performance in on-orbit training of SR and PFL image processing models compared to traditional centralized approaches. This novel method shows significant improvements in data acquisition efficiency, process accuracy, and model adaptability to local satellite conditions.
- 24 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 24 (8) , 6945-6960
ALANINE: A Novel Decentralized Personalized Federated Learning for Heterogeneous LEO Satellite Constellation
Low Earth Orbit (LEO) satellite constellations have seen significant growth and functional enhancement in recent years, which integrates various capabilities like communication, navigation, and remote sensing. However, the heterogeneity of data collected by different satellites and the problems of efficient inter-satellite collaborative computation pose significant obstacles to realizing the potential of these constellations. Existing approaches struggle with data heterogeneity, varing image resolutions, and the need for efficient on-orbit model training. To address these challenges, we propose a novel decentralized PFL framework, namely, A Novel DecentraLized PersonAlized Federated Learning for HeterogeNeous LEO SatellIte CoNstEllation (ALANINE). ALANINE incorporates decentralized FL (DFL) for satellite image Super Resolution (SR), which enhances input data quality. Then it utilizes PFL to implement a personalized approach that accounts for unique characteristics of satellite data. In addition, the framework employs advanced model pruning to optimize model complexity and transmission efficiency. The framework enables efficient data acquisition and processing while improving the accuracy of PFL image processing models. Simulation results demonstrate that ALANINE exhibits superior performance in on-orbit training of SR and PFL image processing models compared to traditional centralized approaches. This novel method shows significant improvements in data acquisition efficiency, process accuracy, and model adaptability to local satellite conditions.