Xiaohui Zhao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Xiaohui Zhao has a total of 101 co-authors. Here are the top co-authors by number of shared publications:
- Zan Li
Jilin University
19 shared publications - Hui Liang
School of Electrical Engineering and Intelligentization, Dongguan University of Technology, Dongguan, China
6 shared publications - Shaoai Guo
College of Communication Engineering, Jilin University, Changchun, Jilin, China
5 shared publications - Lingling Chen
College of Information and Control Engineering, Jilin Institute of Chemical Technology, Jilin 132022, China
4 shared publications - Torsten Braun
Institute of Computer Science, University of Bern, Bern, Switzerland
4 shared publications - Haiyan Wang
College of Communication Engineering, Jilin University, Changchun, China
3 shared publications
Latest Publications
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 22 (11) , 10355-10366
Modulation Recognition on FPGA via Lightweight Transformer With Confidence-Gated Quantization-Aware Distillation
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 22 (11) , 10355-10366
Modulation Recognition on FPGA via Lightweight Transformer With Confidence-Gated Quantization-Aware Distillation
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 22 (7) , 6558-6568
CrowdGraph: Crowdsourcing Positioning Based on Multimodal Graph Attention Model
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 22 (7) , 6558-6568
CrowdGraph: Crowdsourcing Positioning Based on Multimodal Graph Attention Model
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (23) , 50642-50652
A Universal Speech Semantic Communication Framework for Multitask Applications Based on Unsupervised Models
With the increasing complexity of the next-generation network applications and the coexistence of diverse service requirements, generative AI (GAI) and large model (LM)-based semantic communication are widely regarded as promising solutions to address these challenges. The goal of these systems is not only to reduce system burden by reducing transmission data, but also to adapt to new and complex requirements. In this article, we propose a semantic communication system designed to meet the diverse requirements of speech applications while enabling accurate speech transmission. The semantic encoder comprises an unsupervised model wav2vec 2.0 for learning universal speech representations to enable adaptability across various speech-related tasks. It also includes a prosodic feature encoder from the style embedding module of global style token (GST)-Tacotron. The semantic decoder integrates a phoneme recognition module and a GST-Tacotron-based text-to-speech (TTS) module to facilitate accurate and expressive reconstruction of the original speech signal, with the incorporation of prosodic features enhancing the naturalness and intelligibility of the synthesized speech. The proposed system has been tested in noisy channels. It demonstrates that the system maintains superior and robust performance even at bit error rate (BER) of $10^{-1}$ , as reflected by a stable character error rate (CER) approximately 0.0940 and 0.0649 for the base and large versions of wav2vec 2.0, respectively, in speech recognition, and consistent conditional Fréchet deep speech distance (cFDSD) scores approximately 0.9 in speech quality assessment. This performance surpasses that of the existing semantic communication systems, while also providing reliable support for a wider range of downstream speech applications.
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (23) , 50642-50652
A Universal Speech Semantic Communication Framework for Multitask Applications Based on Unsupervised Models
With the increasing complexity of the next-generation network applications and the coexistence of diverse service requirements, generative AI (GAI) and large model (LM)-based semantic communication are widely regarded as promising solutions to address these challenges. The goal of these systems is not only to reduce system burden by reducing transmission data, but also to adapt to new and complex requirements. In this article, we propose a semantic communication system designed to meet the diverse requirements of speech applications while enabling accurate speech transmission. The semantic encoder comprises an unsupervised model wav2vec 2.0 for learning universal speech representations to enable adaptability across various speech-related tasks. It also includes a prosodic feature encoder from the style embedding module of global style token (GST)-Tacotron. The semantic decoder integrates a phoneme recognition module and a GST-Tacotron-based text-to-speech (TTS) module to facilitate accurate and expressive reconstruction of the original speech signal, with the incorporation of prosodic features enhancing the naturalness and intelligibility of the synthesized speech. The proposed system has been tested in noisy channels. It demonstrates that the system maintains superior and robust performance even at bit error rate (BER) of $10^{-1}$ , as reflected by a stable character error rate (CER) approximately 0.0940 and 0.0649 for the base and large versions of wav2vec 2.0, respectively, in speech recognition, and consistent conditional Fréchet deep speech distance (cFDSD) scores approximately 0.9 in speech quality assessment. This performance surpasses that of the existing semantic communication systems, while also providing reliable support for a wider range of downstream speech applications.
- 22 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (21) , 45307-45324
Robust Joint Optimization for Efficient and Reliable FSO/RF Satellite-UAV-Terrestrial Networks With Random Fading and Imperfect Channel Information
This study focuses on a free space optical (FSO)/radio frequency (RF) satellite-UAV-terrestrial integrated network (SUTIN) to overcome limitations of traditional RF SUTIN on spectrum scarcity and security. Traditional resource allocations for FSO/RF SUTIN often prioritize transmission efficiency while neglecting reliability. Moreover, in this network, atmospheric turbulence, pointing errors, and multipath fading can lead to unreliable data transmission. To resolve these problems, we propose a hybrid FSO/RF SUTIN based on adaptive modulation and coding schemes (MCS) and resource allocation algorithm to ensure both transmission efficiency and reliability with imperfect channel state information (CSI) and random channel fading. A joint optimization problem is formulated, incorporating MCS mode selection, power allocation, and UAV trajectory optimization, with frame error rate (FER) and CSI uncertainty constraints to maximize system throughput. To solve this complicated high dimension optimization problem, we present mathematical mapping models between the optimization variables and communication performance parameters (the average transmission rate and FER) with imperfect CSI and random fading for both FSO and RF link. According to these mapping models, a robust solution based on deep deterministic policy gradient is proposed, integrating decoupling and user clustering techniques to reduce the computational complexity. The simulation results demonstrate that this proposed FSO/RF SUTIN and the corresponding algorithm can improve system throughput and guarantee transmission reliability.
- 22 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (21) , 45307-45324
Robust Joint Optimization for Efficient and Reliable FSO/RF Satellite-UAV-Terrestrial Networks With Random Fading and Imperfect Channel Information
This study focuses on a free space optical (FSO)/radio frequency (RF) satellite-UAV-terrestrial integrated network (SUTIN) to overcome limitations of traditional RF SUTIN on spectrum scarcity and security. Traditional resource allocations for FSO/RF SUTIN often prioritize transmission efficiency while neglecting reliability. Moreover, in this network, atmospheric turbulence, pointing errors, and multipath fading can lead to unreliable data transmission. To resolve these problems, we propose a hybrid FSO/RF SUTIN based on adaptive modulation and coding schemes (MCS) and resource allocation algorithm to ensure both transmission efficiency and reliability with imperfect channel state information (CSI) and random channel fading. A joint optimization problem is formulated, incorporating MCS mode selection, power allocation, and UAV trajectory optimization, with frame error rate (FER) and CSI uncertainty constraints to maximize system throughput. To solve this complicated high dimension optimization problem, we present mathematical mapping models between the optimization variables and communication performance parameters (the average transmission rate and FER) with imperfect CSI and random fading for both FSO and RF link. According to these mapping models, a robust solution based on deep deterministic policy gradient is proposed, integrating decoupling and user clustering techniques to reduce the computational complexity. The simulation results demonstrate that this proposed FSO/RF SUTIN and the corresponding algorithm can improve system throughput and guarantee transmission reliability.
- Open Access
- 30 December 2024
- Published by MDPI AG in Electronics
- Vol. 14 (1) , 116
Dual-Layer Path Planning Model for Autonomous Vehicles in Urban Road Networks Using an Improved Deep Q-Network Algorithm with Proportional–Integral–Derivative Control
With the continuous progress of intelligent transportation systems and automated driving technologies, complex urban road environments put forward higher requirements on the real-time characteristic and accuracy of path planning algorithms. Traditional single-layer path planning methods struggle to effectively handle the complexity of road and lane networks, leading to high computational complexity and suboptimal planning outcomes. To address this issue, we propose a dual-layer path planning model. First, at the road level, we employ the A* algorithm to efficiently determine the optimal macroscopic route, reducing computational overhead. At the lane level, we introduce a Proportional–Integral–Derivative Q-network (PIDQN) based on deep reinforcement learning, which leverages PID control mechanisms to enhance lane selection accuracy and adaptability. By incorporating proportional, integral, and derivative control, PIDQN effectively handles dynamic environments and avoids local optima, ensuring stable and faster convergence. Compared with traditional Deep Q-Network (DQN) and Q-learning algorithms, PIDQN demonstrates significant improvements in success rate and convergence speed in path planning tasks. Using high-precision maps in real-world environments and Python for simulation experiments, we verify the superiority of this approach in complex urban road networks, and we compare the performance of traditional A* algorithms and two-layer planning algorithms. The results show that the two-layer planning algorithm outperforms the traditional A* algorithm and provides a more robust and efficient solution for self-driving car navigation.
- 30 December 2024
- Published by MDPI AG in Electronics
- Vol. 14 (1) , 116
Open AccessDual-Layer Path Planning Model for Autonomous Vehicles in Urban Road Networks Using an Improved Deep Q-Network Algorithm with Proportional–Integral–Derivative Control
With the continuous progress of intelligent transportation systems and automated driving technologies, complex urban road environments put forward higher requirements on the real-time characteristic and accuracy of path planning algorithms. Traditional single-layer path planning methods struggle to effectively handle the complexity of road and lane networks, leading to high computational complexity and suboptimal planning outcomes. To address this issue, we propose a dual-layer path planning model. First, at the road level, we employ the A* algorithm to efficiently determine the optimal macroscopic route, reducing computational overhead. At the lane level, we introduce a Proportional–Integral–Derivative Q-network (PIDQN) based on deep reinforcement learning, which leverages PID control mechanisms to enhance lane selection accuracy and adaptability. By incorporating proportional, integral, and derivative control, PIDQN effectively handles dynamic environments and avoids local optima, ensuring stable and faster convergence. Compared with traditional Deep Q-Network (DQN) and Q-learning algorithms, PIDQN demonstrates significant improvements in success rate and convergence speed in path planning tasks. Using high-precision maps in real-world environments and Python for simulation experiments, we verify the superiority of this approach in complex urban road networks, and we compare the performance of traditional A* algorithms and two-layer planning algorithms. The results show that the two-layer planning algorithm outperforms the traditional A* algorithm and provides a more robust and efficient solution for self-driving car navigation.
- 5 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24 (2) , 1221-1233
MR-Transformer: FPGA Accelerated Deep Learning Attention Model for Modulation Recognition
Modulation recognition has emerged as an intensive research topic to improve the communication efficiency in the future 6G network and plays an important role in the security of electromagnetic spectrum. Various pattern recognition methods have been proposed to enhance the performance of modulation recognition, especially deep learning models showing their emerging performance. In this work, we design a modulation recognition model based on an enhanced Transformer, namely MR-Transformer, which is accelerated on a Field Programmable Gate Array (FPGA). The design of MR-Transformer targets on high recognition accuracy, low power consumption, and high computation efficiency, which is suitable for modulation recognition at edge devices. MR-Transformer leverages attention mechanism to extract global features and correspondingly enhance the recognition accuracy. An improved matrix multiplication operation and enhanced Design Space Exploration (DSE) method are proposed in MR-Transformer to improve the computation efficiency and reduce resource consumption. We conduct comprehensive experiments to evaluate the performance of MR-Transformer on three platforms, i.e. Central Processing Unit (CPU), Graphics Processing Unit (GPU), and FPGA, based on two open-source datasets. According to the evaluation results, the MR-Transformer based on FPGA shows the best performance compared with the baseline models considering accuracy, power consumption, and computation efficiency.
- 5 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24 (2) , 1221-1233
MR-Transformer: FPGA Accelerated Deep Learning Attention Model for Modulation Recognition
Modulation recognition has emerged as an intensive research topic to improve the communication efficiency in the future 6G network and plays an important role in the security of electromagnetic spectrum. Various pattern recognition methods have been proposed to enhance the performance of modulation recognition, especially deep learning models showing their emerging performance. In this work, we design a modulation recognition model based on an enhanced Transformer, namely MR-Transformer, which is accelerated on a Field Programmable Gate Array (FPGA). The design of MR-Transformer targets on high recognition accuracy, low power consumption, and high computation efficiency, which is suitable for modulation recognition at edge devices. MR-Transformer leverages attention mechanism to extract global features and correspondingly enhance the recognition accuracy. An improved matrix multiplication operation and enhanced Design Space Exploration (DSE) method are proposed in MR-Transformer to improve the computation efficiency and reduce resource consumption. We conduct comprehensive experiments to evaluate the performance of MR-Transformer on three platforms, i.e. Central Processing Unit (CPU), Graphics Processing Unit (GPU), and FPGA, based on two open-source datasets. According to the evaluation results, the MR-Transformer based on FPGA shows the best performance compared with the baseline models considering accuracy, power consumption, and computation efficiency.
- 28 June 2024
- Published by Springer Nature in Telecommunication Systems
- Vol. 87 (2) , 359-383
A dynamic spectrum access algorithm based on deep reinforcement learning with novel multi-vehicle reward functions in cognitive vehicular networks
- 28 June 2024
- Published by Springer Nature in Telecommunication Systems
- Vol. 87 (2) , 359-383
A dynamic spectrum access algorithm based on deep reinforcement learning with novel multi-vehicle reward functions in cognitive vehicular networks
- 27 March 2024
- Published by Elsevier in Pervasive and Mobile Computing
- Vol. 101
Defending dominant cooperative probabilistic attack in CRNs by JS-divergence-based improved reputation algorithm
- 27 March 2024
- Published by Elsevier in Pervasive and Mobile Computing
- Vol. 101
Defending dominant cooperative probabilistic attack in CRNs by JS-divergence-based improved reputation algorithm
- Top Cited Papers
- 27 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (11) , 19882-19897
Deep-Reinforcement-Learning-Based Computation Offloading in UAV-Assisted Vehicular Edge Computing Networks
Vehicular edge computing (VEC) is considered to be a key technology to improve the processing efficiency of computing tasks for the Internet of Vehicles (IoV). Using roadside units (RSUs) distributed on both sides of a road as edge servers, computation-intensive and latency-sensitive in-vehicle tasks can be responded to quickly. However, some Quality of Service (QoS) is often difficult to ensure due to clogged dense urban buildings or lack of infrastructure in remote areas. In this article, we propose a software-defined network (SDN)-driven partial offloading model for unmanned aerial vehicle (UAV)-assisted VEC networks, where the RSUs and UAVs jointly provide computing services to the vehicles and collect global information through centralized control using an SDN controller. To guarantee these vehicles obtain computing results in time and rationally utilize computing resources, we develop an optimal offloading mechanism using Age of Information (AoI), together with energy consumption and rental price as a comprehensive weighted cost of our above optimization objective. The total system cost of the performing tasks is minimized by jointly optimizing the UAV trajectory, user association, and offloading decision. Considering the mobility of the vehicles and UAVs and the dynamic network environment, we design a deep reinforcement learning (DRL)-based joint trajectory control and offloading allocation algorithm (DRL-TCOA) to solve the proposed computation offloading problem. Experimental results show that the proposed DRL-TCOA algorithm maintains better information freshness and lower system cost than the other baseline offloading strategies.
- 27 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (11) , 19882-19897
Top Cited PapersDeep-Reinforcement-Learning-Based Computation Offloading in UAV-Assisted Vehicular Edge Computing Networks
Vehicular edge computing (VEC) is considered to be a key technology to improve the processing efficiency of computing tasks for the Internet of Vehicles (IoV). Using roadside units (RSUs) distributed on both sides of a road as edge servers, computation-intensive and latency-sensitive in-vehicle tasks can be responded to quickly. However, some Quality of Service (QoS) is often difficult to ensure due to clogged dense urban buildings or lack of infrastructure in remote areas. In this article, we propose a software-defined network (SDN)-driven partial offloading model for unmanned aerial vehicle (UAV)-assisted VEC networks, where the RSUs and UAVs jointly provide computing services to the vehicles and collect global information through centralized control using an SDN controller. To guarantee these vehicles obtain computing results in time and rationally utilize computing resources, we develop an optimal offloading mechanism using Age of Information (AoI), together with energy consumption and rental price as a comprehensive weighted cost of our above optimization objective. The total system cost of the performing tasks is minimized by jointly optimizing the UAV trajectory, user association, and offloading decision. Considering the mobility of the vehicles and UAVs and the dynamic network environment, we design a deep reinforcement learning (DRL)-based joint trajectory control and offloading allocation algorithm (DRL-TCOA) to solve the proposed computation offloading problem. Experimental results show that the proposed DRL-TCOA algorithm maintains better information freshness and lower system cost than the other baseline offloading strategies.