Shizhen Xiao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Shizhen Xiao has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Jing Mei
College of Information Science and Engineering, Hunan Normal University, Changsha, Hunan, China
2 shared publications - Keqin Li
Hunan University
2 shared publications Lihui Xia
College of Information Science and Engineering, Hunan Normal University in Changsha, Hunan, China
2 shared publications- Zhao Tong
College of Information Science and Engineering, Hunan Normal University, Changsha, Hunan, China
2 shared publications Xi Zhang
College of Information Science and Engineering, Hunan Normal University in Changsha, Hunan, China
1 shared publication- Xi Zhang
College of Information Science and Engineering, Hunan Normal University, Changsha, Hunan, China
1 shared publication
Latest Publications
- 26 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. PP (15361233) , 1-17
Transformer Enhanced Task Offloading with Privacy-Preserving in Heterogeneous and Dynamic Edge Networks
The exponential growth of edge devices and growing demand for low-latency, high-throughput applications have established edge computing as essential infrastructure. Edge computing enables near-data computation to alleviate device burdens, yet task offloading optimization remains challenging due to the heterogeneous, dynamic networks and privacy risks in centralized approaches. To address these intertwined challenges, we propose the federated Q-Learning via Transformer for task offloading (FQTTO) algorithm, which integrates federated learning (FL) with a Transformer-based Q-Network. FL ensures privacy-preserving distributed model updates, while the Transformer employs customized sequential state encodings of task priorities and resource availabilities to autoregressively predict Q-values, incorporating $n$-step returns for long-term optimization. Extensive simulations show that compared to the baseline methods, FQTTO achieves average reductions exceeding 20.4% in delay and 22.3% in energy consumption, while improving load balance by at least 13.4% and enhancing the task completion rate by up to 12.1%.
- 26 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. PP (15361233) , 1-17
Transformer Enhanced Task Offloading with Privacy-Preserving in Heterogeneous and Dynamic Edge Networks
The exponential growth of edge devices and growing demand for low-latency, high-throughput applications have established edge computing as essential infrastructure. Edge computing enables near-data computation to alleviate device burdens, yet task offloading optimization remains challenging due to the heterogeneous, dynamic networks and privacy risks in centralized approaches. To address these intertwined challenges, we propose the federated Q-Learning via Transformer for task offloading (FQTTO) algorithm, which integrates federated learning (FL) with a Transformer-based Q-Network. FL ensures privacy-preserving distributed model updates, while the Transformer employs customized sequential state encodings of task priorities and resource availabilities to autoregressively predict Q-values, incorporating $n$-step returns for long-term optimization. Extensive simulations show that compared to the baseline methods, FQTTO achieves average reductions exceeding 20.4% in delay and 22.3% in energy consumption, while improving load balance by at least 13.4% and enhancing the task completion rate by up to 12.1%.
- 16 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (12) , 25980-25995
LADPG: A Lyapunov Optimization-Based Resource Allocation Strategy for Heterogeneous Vehicular Networks
Vehicular edge computing (VEC) is an emerging paradigm that offloads computationally intensive tasks to nearby VEC servers instead of cloud servers, effectively reducing data transmission latency and enhancing system real-time performance and reliability. However, heterogeneous communication technologies and task diversity introduce significant challenges in resource allocation, rendering traditional optimization methods inadequate for dynamic network environments and long-term queue stability requirements. This article proposed a Lyapunov adaptive deterministic policy gradient (LADPG) algorithm to address the dynamic multiobjective optimization problem of online task offloading and resource allocation in a single VEC server scenario. LADPG integrates Lyapunov optimization with the deep deterministic policy gradient (DDPG) framework to transform long-term queue stability constraints into short-term reward functions. It dynamically adjusts communication and VEC server computation resource allocations to jointly optimize energy consumption and utility costs under latency constraints. In addition, a dynamic channel selection mechanism is designed to mitigate congestion and enhance transmission efficiency. Extensive simulations demonstrate that LADPG significantly improves system performance, particularly in high-load and dynamic scenarios.
- 16 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (12) , 25980-25995
LADPG: A Lyapunov Optimization-Based Resource Allocation Strategy for Heterogeneous Vehicular Networks
Vehicular edge computing (VEC) is an emerging paradigm that offloads computationally intensive tasks to nearby VEC servers instead of cloud servers, effectively reducing data transmission latency and enhancing system real-time performance and reliability. However, heterogeneous communication technologies and task diversity introduce significant challenges in resource allocation, rendering traditional optimization methods inadequate for dynamic network environments and long-term queue stability requirements. This article proposed a Lyapunov adaptive deterministic policy gradient (LADPG) algorithm to address the dynamic multiobjective optimization problem of online task offloading and resource allocation in a single VEC server scenario. LADPG integrates Lyapunov optimization with the deep deterministic policy gradient (DDPG) framework to transform long-term queue stability constraints into short-term reward functions. It dynamically adjusts communication and VEC server computation resource allocations to jointly optimize energy consumption and utility costs under latency constraints. In addition, a dynamic channel selection mechanism is designed to mitigate congestion and enhance transmission efficiency. Extensive simulations demonstrate that LADPG significantly improves system performance, particularly in high-load and dynamic scenarios.