Hao Hu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Hao Hu has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Zhaolong Ning
School of Communications and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing, China
6 shared publications - Xiaojie Wang
School of Communications and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing, China
4 shared publications - Yan Zhang
University of Oslo
3 shared publications - Chau Yuen
School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore, Singapore
2 shared publications - Amr Tolba
King Saud University
1 shared publication - Fan Wu
School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China
1 shared publication
Latest Publications
- 30 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. PP (15361233) , 1-15
Dynamically Segmented IRS-Assisted UAV Computing Power Networks: Towards System Delay and Energy Consumption Optimization
In this paper, we propose a dynamically segmented Intelligent Reflecting Surface (IRS)-assisted Unmanned Aerial Vehicle (UAV) Computing Power Networks (CPNs) with tightly integrated communication and computing power resources. The IRS can be dynamically segmented and allocated to users, with computing resources allocated accordingly to satisfy their delay constraints. Considering the energy limitations of UAVs, we formulate a multi-objective optimization problem to minimize user delay and UAV energy consumption. To solve the problem, we propose a new scheme jointly considering UAV trajectory, computing power allocation, reflecting element allocation, phase shift, and UAV-user association (TCPA) scheme. The phase alignment theory is utilized to determine the IRS phase shift control and decompose the problem into three subproblems based on the coupling of variables. Specifically, we use channel optimal matching to solve the first subproblem to obtain user association decisions. Then, we formulate a computing power communication matching subproblem, and propose a successive convex approximation scheme to solve it. The trajectory subproblem is optimized by a multi-agent deep reinforcement learning-based method. The evaluation results demonstrate that our proposed TCPA achieves high performance in terms of reward and system delay. Additionally, it demonstrates that integrating IRS and CPNs can effectively reduce the total system delay with only a marginal increase in energy consumption.
- 30 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. PP (15361233) , 1-15
Dynamically Segmented IRS-Assisted UAV Computing Power Networks: Towards System Delay and Energy Consumption Optimization
In this paper, we propose a dynamically segmented Intelligent Reflecting Surface (IRS)-assisted Unmanned Aerial Vehicle (UAV) Computing Power Networks (CPNs) with tightly integrated communication and computing power resources. The IRS can be dynamically segmented and allocated to users, with computing resources allocated accordingly to satisfy their delay constraints. Considering the energy limitations of UAVs, we formulate a multi-objective optimization problem to minimize user delay and UAV energy consumption. To solve the problem, we propose a new scheme jointly considering UAV trajectory, computing power allocation, reflecting element allocation, phase shift, and UAV-user association (TCPA) scheme. The phase alignment theory is utilized to determine the IRS phase shift control and decompose the problem into three subproblems based on the coupling of variables. Specifically, we use channel optimal matching to solve the first subproblem to obtain user association decisions. Then, we formulate a computing power communication matching subproblem, and propose a successive convex approximation scheme to solve it. The trajectory subproblem is optimized by a multi-agent deep reinforcement learning-based method. The evaluation results demonstrate that our proposed TCPA achieves high performance in terms of reward and system delay. Additionally, it demonstrates that integrating IRS and CPNs can effectively reduce the total system delay with only a marginal increase in energy consumption.
- 24 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- Vol. 40 (3) , 67-74
AI Agent for Task Assignment in End-Edge-Cloud Networks
Artificial Intelligence (AI) agents empower end-edge-cloud networks with intelligent interaction and decision-making capabilities, in which Large Language Models (LLMs) provide strong global information processing and lightweight models offer rapid responses. The integration of diverse AI agents into networks enables the exploitation of global insights for strategic decisions. However, the heterogeneous devices in end-edge-cloud networks pose significant challenges for efficient task assignment. In addition, the redundancy of global information and the limitations of local information further complicate collaborative decision-making. To this end, we propose an end-edge-cloud network with AI agents, in which LLMs and lightweight models are deployed at different layers to enable collaborative decision-making. The collaboration of diverse AI agents across multiple layers enables rapid local responses while ensuring effective extraction of global features, thereby enhancing task assignment efficiency in end-edge-cloud networks. Simulation results indicate that the collaboration between LLMs and lightweight models facilitates efficient task assignment in end-edge-cloud networks, significantly increasing the total number of completed tasks.
- 24 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Network
- Vol. 40 (3) , 67-74
AI Agent for Task Assignment in End-Edge-Cloud Networks
Artificial Intelligence (AI) agents empower end-edge-cloud networks with intelligent interaction and decision-making capabilities, in which Large Language Models (LLMs) provide strong global information processing and lightweight models offer rapid responses. The integration of diverse AI agents into networks enables the exploitation of global insights for strategic decisions. However, the heterogeneous devices in end-edge-cloud networks pose significant challenges for efficient task assignment. In addition, the redundancy of global information and the limitations of local information further complicate collaborative decision-making. To this end, we propose an end-edge-cloud network with AI agents, in which LLMs and lightweight models are deployed at different layers to enable collaborative decision-making. The collaboration of diverse AI agents across multiple layers enables rapid local responses while ensuring effective extraction of global features, thereby enhancing task assignment efficiency in end-edge-cloud networks. Simulation results indicate that the collaboration between LLMs and lightweight models facilitates efficient task assignment in end-edge-cloud networks, significantly increasing the total number of completed tasks.
- 9 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25 (15361276) , 8483-8497
Channel-Aware User Association and Trajectory Design for Multi-IRS Assisted Multi-UAV Communications
The integration of Intelligent Reflecting Surfaces (IRSs) and Unmanned Aerial Vehicles (UAVs) is promising for providing flexible and intelligent communications to users in urban areas. Existing studies are founded either on the complete Line of Sight (LoS) or complete Non-LoS (NLoS) communication scenarios, while ignoring their coexistence. To solve the above challenge in complicated and dynamic communication scenarios, we formulate an average system sum rate maximization problem with the optimization of joint IRS-user association, multi-UAV trajectory optimization, IRS phase shifts and transmit power allocation. Since the highly complex and coupled variables, we propose a Multi-Agent Deep Reinforcement Learning (MADRL)-based scheme to maximize the average system sum rate. First, we derive two composite channel power gains for different communication conditions. Then, phase alignment theory is utilized to obtain optimal phase control. To guarantee long-term optimization, we propose a scheme based on Multi-Agent Proximal Policy Optimization (MAPPO) and Successive Convex Approximation (SCA) method to jointly optimize multi-UAV trajectories, multi-IRS association and transmit power allocation. Finally, experimental results reveal that the proposed MGBA shows considerable advantages in both the convergence speed and the average system sum rate.
- 9 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25 (15361276) , 8483-8497
Channel-Aware User Association and Trajectory Design for Multi-IRS Assisted Multi-UAV Communications
The integration of Intelligent Reflecting Surfaces (IRSs) and Unmanned Aerial Vehicles (UAVs) is promising for providing flexible and intelligent communications to users in urban areas. Existing studies are founded either on the complete Line of Sight (LoS) or complete Non-LoS (NLoS) communication scenarios, while ignoring their coexistence. To solve the above challenge in complicated and dynamic communication scenarios, we formulate an average system sum rate maximization problem with the optimization of joint IRS-user association, multi-UAV trajectory optimization, IRS phase shifts and transmit power allocation. Since the highly complex and coupled variables, we propose a Multi-Agent Deep Reinforcement Learning (MADRL)-based scheme to maximize the average system sum rate. First, we derive two composite channel power gains for different communication conditions. Then, phase alignment theory is utilized to obtain optimal phase control. To guarantee long-term optimization, we propose a scheme based on Multi-Agent Proximal Policy Optimization (MAPPO) and Successive Convex Approximation (SCA) method to jointly optimize multi-UAV trajectories, multi-IRS association and transmit power allocation. Finally, experimental results reveal that the proposed MGBA shows considerable advantages in both the convergence speed and the average system sum rate.
- 26 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (24) , 39114-39126
Joint Resource Allocation and Trajectory Optimization for Reliable UAV-to-Vehicle Services
Ground-air cooperative package distribution is a promising delivery method, especially during Corona virus Disease 2019. It can extend the coverage of vehicles by exploring the flexibility of unmanned aerial vehicles (UAVs), expand the distribution of vehicles and reduce carbon emissions. Most existing studies focus on their trajectory optimization, while often overlooking their coordination for global information, and the complexity and reliability of collaborative delivery problem. To address the above issues, we first formulate an optimization problem to minimize the service cost of both UAVs and vehicles. To ensure service reliability, the constraints of UAVs during takeoff, service, and landing phases are comprehensively considered. We then propose a lightweight reinforcement learning solution to minimize the flight distance of UAVs and the number of required vehicles. Finally, theoretical analysis and performance evaluations show that compared with other representative algorithms, the designed algorithm has advantages in terms of robustness, effectiveness and stability.
- 26 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (24) , 39114-39126
Joint Resource Allocation and Trajectory Optimization for Reliable UAV-to-Vehicle Services
Ground-air cooperative package distribution is a promising delivery method, especially during Corona virus Disease 2019. It can extend the coverage of vehicles by exploring the flexibility of unmanned aerial vehicles (UAVs), expand the distribution of vehicles and reduce carbon emissions. Most existing studies focus on their trajectory optimization, while often overlooking their coordination for global information, and the complexity and reliability of collaborative delivery problem. To address the above issues, we first formulate an optimization problem to minimize the service cost of both UAVs and vehicles. To ensure service reliability, the constraints of UAVs during takeoff, service, and landing phases are comprehensively considered. We then propose a lightweight reinforcement learning solution to minimize the flight distance of UAVs and the number of required vehicles. Finally, theoretical analysis and performance evaluations show that compared with other representative algorithms, the designed algorithm has advantages in terms of robustness, effectiveness and stability.
- 22 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23 (10) , 13408-13423
Joint User Association, Interference Cancellation, and Power Control for Multi-IRS Assisted UAV Communications
Intelligent reflecting surface (IRS)-assisted unmanned aerial vehicle (UAV) communications are expected to alleviate the load of ground base stations in a cost-effective way. Existing studies mainly focus on the deployment and resource allocation of a single IRS instead of multiple IRSs, whereas it is extremely challenging for joint multi-IRS multi-user association in UAV communications with constrained reflecting resources and dynamic scenarios. To address the aforementioned challenges, we propose a new optimization algorithm for joint IRS-user association, trajectory optimization of UAVs, successive interference cancellation (SIC) decoding order scheduling and power allocation to maximize system energy efficiency. We first propose an inverse soft-Q learning-based algorithm to optimize multi-IRS multi-user association. Then, successive convex approximation (SCA) and Dinkelbach-based algorithm are leveraged to optimize UAV trajectory followed by the optimization of SIC decoding order scheduling and power allocation. Finally, theoretical analysis and performance results show significant advantages of the designed algorithm in convergence rate and energy efficiency.
- 22 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23 (10) , 13408-13423
Joint User Association, Interference Cancellation, and Power Control for Multi-IRS Assisted UAV Communications
Intelligent reflecting surface (IRS)-assisted unmanned aerial vehicle (UAV) communications are expected to alleviate the load of ground base stations in a cost-effective way. Existing studies mainly focus on the deployment and resource allocation of a single IRS instead of multiple IRSs, whereas it is extremely challenging for joint multi-IRS multi-user association in UAV communications with constrained reflecting resources and dynamic scenarios. To address the aforementioned challenges, we propose a new optimization algorithm for joint IRS-user association, trajectory optimization of UAVs, successive interference cancellation (SIC) decoding order scheduling and power allocation to maximize system energy efficiency. We first propose an inverse soft-Q learning-based algorithm to optimize multi-IRS multi-user association. Then, successive convex approximation (SCA) and Dinkelbach-based algorithm are leveraged to optimize UAV trajectory followed by the optimization of SIC decoding order scheduling and power allocation. Finally, theoretical analysis and performance results show significant advantages of the designed algorithm in convergence rate and energy efficiency.
- 26 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (1) , 3392-3402
A Depression Detection Auxiliary Decision System Based on Multi-Modal Feature-Level Fusion of EEG and Speech
By improving the accuracy of depression recognition and designing a consumer-oriented depression detection system, consumers are expected to receive convenient and fast e-health services. Recently, depression recognition methods based on the analysis of physiological and behavioral data have attracted attention. In particular, the research on Electroencephalography (EEG) and speech signals becomes hotspots. However, EEG is susceptible to individual differences, while speech signal is susceptible to environmental factors. In this study, we propose an auxiliary decision-making system for depression detection that considers both physiological and behavioral factors by fusing EEG and speech signals. Compared to existing studies, our proposed multi-modal fusion strategy exploits more linear and nonlinear features to support the recognition of task classifications. In addition, we analyze the functional connectivity of brain regions to facilitate EEG feature extraction. Considering the non-stationary feature of EEG and speech signals, we perform filtering, artifact processing, and time-frequency domain processing. Furthermore, we integrate the EEG and speech signals at the feature level and train their classification. Performance evaluation results show that our proposed multi-modal feature fusion strategy achieves 86.11% accuracy on the dataset of major depressive disorders, and 87.44% recognition accuracy on the healthy controls.
- 26 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (1) , 3392-3402
A Depression Detection Auxiliary Decision System Based on Multi-Modal Feature-Level Fusion of EEG and Speech
By improving the accuracy of depression recognition and designing a consumer-oriented depression detection system, consumers are expected to receive convenient and fast e-health services. Recently, depression recognition methods based on the analysis of physiological and behavioral data have attracted attention. In particular, the research on Electroencephalography (EEG) and speech signals becomes hotspots. However, EEG is susceptible to individual differences, while speech signal is susceptible to environmental factors. In this study, we propose an auxiliary decision-making system for depression detection that considers both physiological and behavioral factors by fusing EEG and speech signals. Compared to existing studies, our proposed multi-modal fusion strategy exploits more linear and nonlinear features to support the recognition of task classifications. In addition, we analyze the functional connectivity of brain regions to facilitate EEG feature extraction. Considering the non-stationary feature of EEG and speech signals, we perform filtering, artifact processing, and time-frequency domain processing. Furthermore, we integrate the EEG and speech signals at the feature level and train their classification. Performance evaluation results show that our proposed multi-modal feature fusion strategy achieves 86.11% accuracy on the dataset of major depressive disorders, and 87.44% recognition accuracy on the healthy controls.
- 30 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (12) , 21299-21311
Joint UAV Deployment and User Scheduling for Wireless Powered Wearable Networks
The integration of unmanned aerial vehicles (UAVs) and wireless power transfer (WPT) is promising to provide charging and communication services to on-ground devices. However, traditional studies focus on UAVs at a fixed height to provide services for devices, ignoring their flexibility. To fully exploit the potential of UAVs in wearable networks, we take the height of UAVs into consideration to maximize the network throughput and minimize the task completion time. Considering the UAV coverage and energy constraints, we formulate the service time allocation and UAV height control optimization problem as a constrained Markov decision process (CMDP). Then, we propose a Lagrangian-based proximal policy optimization (PPO)-CMDP algorithm to optimize the primal pair-based policy. Finally, simulation evaluation shows that our proposed algorithm has significant advantages in terms of network throughput and service fairness.
- 30 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (12) , 21299-21311
Joint UAV Deployment and User Scheduling for Wireless Powered Wearable Networks
The integration of unmanned aerial vehicles (UAVs) and wireless power transfer (WPT) is promising to provide charging and communication services to on-ground devices. However, traditional studies focus on UAVs at a fixed height to provide services for devices, ignoring their flexibility. To fully exploit the potential of UAVs in wearable networks, we take the height of UAVs into consideration to maximize the network throughput and minimize the task completion time. Considering the UAV coverage and energy constraints, we formulate the service time allocation and UAV height control optimization problem as a constrained Markov decision process (CMDP). Then, we propose a Lagrangian-based proximal policy optimization (PPO)-CMDP algorithm to optimize the primal pair-based policy. Finally, simulation evaluation shows that our proposed algorithm has significant advantages in terms of network throughput and service fairness.