Ye 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
Ye Hu has a total of 44 co-authors. Here are the top co-authors by number of shared publications:
- Mingzhe Chen
University of Miami, Miami, FL, USA
12 shared publications - Walid Saad
Department of Electrical and Computer Engineering, Virginia Tech, Alexandria, VA, USA
9 shared publications - Changchuan Yin
Beijing University of Posts and Telecommunications
3 shared publications - Yuchen Liu
North Carolina State University, Raleigh, NC, USA
3 shared publications - Yujiao Zhu
Beijing Laboratory of Advanced Information Network, Beijing University of Posts and Telecommunications, Beijing, China
3 shared publications - Zhaohui Yang
Zhejiang University
3 shared publications
Latest Publications
- 1 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15  (21622337) , 2914-2918
Compressive Beam-Pattern-Aware Near-Field Beam Training via Total Variation Denoising
Extremely large antenna arrays envisioned for 6G incurs near-field effect, where steering vector depends on angles and range simultaneously. Polar-domain near-field codebooks can focus energy accurately but incur extra two-dimensional sweeping overhead; compressed-sensing (CS) approaches with Gaussian-masked DFT sensing offer a lower-overhead alternative. This letter revisits near-field beam training using conventional DFT codebooks. Unlike far-field responses that concentrate energy on a few isolated DFT beams, near-field responses produce contiguous, plateau-like energy segments with sharp transitions in the DFT beamspace. Pure LASSO denoising, therefore, tends to over-shrink magnitudes and fragment plateaus. We propose a beam-pattern–preserving beam training scheme for multiple-path scenarios that combines LASSO with a lightweight denoising pipeline: LASSO to suppress small-amplitude noise, followed by total variation (TV) to maintain plateau levels and edge sharpness. The two proximal steps require no near-field codebook design. Simulations with Gaussian pilots show consistent NMSE and cosine-similarity gains over least squares and LASSO at the same pilot budget.
- 1 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15  (21622337) , 2914-2918
Compressive Beam-Pattern-Aware Near-Field Beam Training via Total Variation Denoising
Extremely large antenna arrays envisioned for 6G incurs near-field effect, where steering vector depends on angles and range simultaneously. Polar-domain near-field codebooks can focus energy accurately but incur extra two-dimensional sweeping overhead; compressed-sensing (CS) approaches with Gaussian-masked DFT sensing offer a lower-overhead alternative. This letter revisits near-field beam training using conventional DFT codebooks. Unlike far-field responses that concentrate energy on a few isolated DFT beams, near-field responses produce contiguous, plateau-like energy segments with sharp transitions in the DFT beamspace. Pure LASSO denoising, therefore, tends to over-shrink magnitudes and fragment plateaus. We propose a beam-pattern–preserving beam training scheme for multiple-path scenarios that combines LASSO with a lightweight denoising pipeline: LASSO to suppress small-amplitude noise, followed by total variation (TV) to maintain plateau levels and edge sharpness. The two proximal steps require no near-field codebook design. Simulations with Gaussian pilots show consistent NMSE and cosine-similarity gains over least squares and LASSO at the same pilot budget.
- 16 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25  (15361276) , 11980-11997
Resource Allocation for Heterogeneous Services in Satellite-Terrestrial IoT Networks With Multi-Access Edge Computing
To address the challenges of Internet of Things (IoT) device diversity and media service heterogeneity in human and machine-type communications, a predominant approach in sixth-generation (6G) networks and beyond is to serve diversified IoT devices by differentiated services. In this paper, a satellite-terrestrial IoT framework with multi-access edge computing (MEC) is investigated for two types of heterogeneous services, data-intensive and computation-intensive service. In our proposed framework, MEC and millimeter wave (mmWave) communication are jointly considered to optimize data- and computation-intensive services, guaranteeing the rate, delay and energy requirements of diversified IoT devices. From the viewpoint of heterogeneous services, we formulate a joint resource allocation problem, in which quality of experience (QoE) of diversified IoT devices are recognized as system utility. Specifically, service offloading, power allocation and computation resource allocation are jointly considered. Since the optimized problem is nonconvex, necessary problem reformulations are conducted to transfer the original problem to convex problems. Furthermore, an alternating iterative method based on deep reinforcement learning (DRL) and CVX technique is adopted to obtain the sub-optimal solution with low computation complexity. Finally, extensive simulations are conducted with different system parameter configurations to verify the effectiveness of our proposed scheme.
- 16 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25  (15361276) , 11980-11997
Resource Allocation for Heterogeneous Services in Satellite-Terrestrial IoT Networks With Multi-Access Edge Computing
To address the challenges of Internet of Things (IoT) device diversity and media service heterogeneity in human and machine-type communications, a predominant approach in sixth-generation (6G) networks and beyond is to serve diversified IoT devices by differentiated services. In this paper, a satellite-terrestrial IoT framework with multi-access edge computing (MEC) is investigated for two types of heterogeneous services, data-intensive and computation-intensive service. In our proposed framework, MEC and millimeter wave (mmWave) communication are jointly considered to optimize data- and computation-intensive services, guaranteeing the rate, delay and energy requirements of diversified IoT devices. From the viewpoint of heterogeneous services, we formulate a joint resource allocation problem, in which quality of experience (QoE) of diversified IoT devices are recognized as system utility. Specifically, service offloading, power allocation and computation resource allocation are jointly considered. Since the optimized problem is nonconvex, necessary problem reformulations are conducted to transfer the original problem to convex problems. Furthermore, an alternating iterative method based on deep reinforcement learning (DRL) and CVX technique is adopted to obtain the sub-optimal solution with low computation complexity. Finally, extensive simulations are conducted with different system parameter configurations to verify the effectiveness of our proposed scheme.
- 11 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25  (15361276) , 6442-6459
Resilient Vehicular Communications Under Imperfect Channel State Information
Cellular vehicle-to-everything (C-V2X) networks provide a promising solution to improve road safety and traffic efficiency. One key challenge in such systems lies in meeting quality-of-service (QoS) requirements of vehicular communication links given limited network resources, particularly under imperfect channel state information (CSI) conditions caused by the highly dynamic environment. In this paper, a novel two-phase framework is proposed to instill resilience into C-V2X networks under unknown imperfect CSI. The resilience of the C-V2X network is defined, quantified, and optimized for the first time through two principal dimensions: absorption phase and adaptation phase. Specifically, the probability distribution function (PDF) of the imperfect CSI is estimated during the absorption phase through dedicated absorption power scheme and resource block (RB) assignment. The estimated PDF is further used to analyze the interplay and reveal the tradeoff between these two phases. Then, a novel metric named hazard rate (HR) is exploited to balance the C-V2X network’s prioritization on absorption and adaptation. Finally, the estimated PDF is exploited in the adaptation phase to recover the network’s QoS through a real-time power allocation optimization. Simulation results demonstrate the superior capability of the proposed framework in sustaining the QoS of the C-V2X network under imperfect CSI. Specifically, in the adaptation phase, the proposed design reduces the vehicle-to-vehicle (V2V) delay that exceeds QoS requirement by 23% and 46%, and improves the average vehicle-to-infrastructure (V2I) throughput by 14% and 16% compared to the model-based and data-driven benchmarks, respectively, without compromising the network’s QoS in the absorption phase.
- 11 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25  (15361276) , 6442-6459
Resilient Vehicular Communications Under Imperfect Channel State Information
Cellular vehicle-to-everything (C-V2X) networks provide a promising solution to improve road safety and traffic efficiency. One key challenge in such systems lies in meeting quality-of-service (QoS) requirements of vehicular communication links given limited network resources, particularly under imperfect channel state information (CSI) conditions caused by the highly dynamic environment. In this paper, a novel two-phase framework is proposed to instill resilience into C-V2X networks under unknown imperfect CSI. The resilience of the C-V2X network is defined, quantified, and optimized for the first time through two principal dimensions: absorption phase and adaptation phase. Specifically, the probability distribution function (PDF) of the imperfect CSI is estimated during the absorption phase through dedicated absorption power scheme and resource block (RB) assignment. The estimated PDF is further used to analyze the interplay and reveal the tradeoff between these two phases. Then, a novel metric named hazard rate (HR) is exploited to balance the C-V2X network’s prioritization on absorption and adaptation. Finally, the estimated PDF is exploited in the adaptation phase to recover the network’s QoS through a real-time power allocation optimization. Simulation results demonstrate the superior capability of the proposed framework in sustaining the QoS of the C-V2X network under imperfect CSI. Specifically, in the adaptation phase, the proposed design reduces the vehicle-to-vehicle (V2V) delay that exceeds QoS requirement by 23% and 46%, and improves the average vehicle-to-infrastructure (V2I) throughput by 14% and 16% compared to the model-based and data-driven benchmarks, respectively, without compromising the network’s QoS in the absorption phase.
- 28 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25  (15361276) , 1408-1421
Optimization of Private Semantic Communication Performance: An Uncooperative Covert Communication Method
In this paper, a novel covert semantic communication framework is investigated. Within this framework, a server extracts and transmits the semantic information, i.e., the meaning of image data, to a user over several time slots. An attacker seeks to detect and eavesdrop the semantic transmission to acquire details of the original image. To avoid data meaning being eavesdropped by an attacker, a friendly jammer is deployed to transmit jamming signals to interfere the attacker so as to hide the transmitted semantic information. Meanwhile, the server will strategically select time slots for semantic information transmission. Due to limited energy, the jammer will not communicate with the server and hence the server does not know the transmit power of the jammer. Therefore, the server must jointly optimize the semantic information transmitted at each time slot and the corresponding transmit power to maximize the privacy and the semantic information transmission quality of the user. To solve this problem, we propose a prioritised sampling assisted twin delayed deep deterministic policy gradient algorithm to jointly determine the transmitted semantic information and the transmit power per time slot without the communications between the server and the jammer. Compared to standard reinforcement learning methods, the proposed method uses an additional Q network to estimate Q values such that the agent can select the action with a lower Q value from the two Q networks thus avoiding local optimal action selection and estimation bias of Q values. Simulation results show that the proposed algorithm can improve the privacy and the semantic information transmission quality by up to 77.8% and 14.3% compared to the traditional reinforcement learning methods.
- 28 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 25  (15361276) , 1408-1421
Optimization of Private Semantic Communication Performance: An Uncooperative Covert Communication Method
In this paper, a novel covert semantic communication framework is investigated. Within this framework, a server extracts and transmits the semantic information, i.e., the meaning of image data, to a user over several time slots. An attacker seeks to detect and eavesdrop the semantic transmission to acquire details of the original image. To avoid data meaning being eavesdropped by an attacker, a friendly jammer is deployed to transmit jamming signals to interfere the attacker so as to hide the transmitted semantic information. Meanwhile, the server will strategically select time slots for semantic information transmission. Due to limited energy, the jammer will not communicate with the server and hence the server does not know the transmit power of the jammer. Therefore, the server must jointly optimize the semantic information transmitted at each time slot and the corresponding transmit power to maximize the privacy and the semantic information transmission quality of the user. To solve this problem, we propose a prioritised sampling assisted twin delayed deep deterministic policy gradient algorithm to jointly determine the transmitted semantic information and the transmit power per time slot without the communications between the server and the jammer. Compared to standard reinforcement learning methods, the proposed method uses an additional Q network to estimate Q values such that the agent can select the action with a lower Q value from the two Q networks thus avoiding local optimal action selection and estimation bias of Q values. Simulation results show that the proposed algorithm can improve the privacy and the semantic information transmission quality by up to 77.8% and 14.3% compared to the traditional reinforcement learning methods.
- 5 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73  (11) , 11123-11137
Performance Optimization of Semantic Communications With Heterogeneous Knowledge: An Adversarial Reinforcement Learning Approach
In this paper, a semantic communication framework where the transmitter and the receiver possess different knowledge (i.e., different methods to extract semantic information and regenerate source data) is investigated. In the proposed framework, the transmitter extracts semantic information according to its knowledge, and the receiver processes the received semantic information based on its own knowledge. To ensure the receiver can understand the semantic information as anticipated, the transmitter will ask a series of questions to estimate the receiver’s knowledge and adjust the method of semantic information extraction according to the estimation. Due to the limited wireless resources and communication time, the size of the extracted semantic information and the number of questions that the transmitter can ask are limited. This problem is formulated as an optimization problem whose goal is to maximize worse case answer similarities over semantic generation and question selection decisions. More specifically, the transmitter aims to ask the questions to pinpoint the largest knowledge divergence and adjusts its semantic generation method to minimize this divergence. An adversarial reinforcement learning (ARL) inspired algorithm, combined with a matching network, is designed to achieve such opposite goals by searching the optimal semantic information generation scheme and question selection scheme in an adversarial manner. Simulation results demonstrate that the proposed framework can improve the semantic similarity of answers by up to 5.5% gain and can achieve up to 10.7% gain in terms of the average similarity of texts compared to the algorithm without knowledge estimation.
- 5 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73  (11) , 11123-11137
Performance Optimization of Semantic Communications With Heterogeneous Knowledge: An Adversarial Reinforcement Learning Approach
In this paper, a semantic communication framework where the transmitter and the receiver possess different knowledge (i.e., different methods to extract semantic information and regenerate source data) is investigated. In the proposed framework, the transmitter extracts semantic information according to its knowledge, and the receiver processes the received semantic information based on its own knowledge. To ensure the receiver can understand the semantic information as anticipated, the transmitter will ask a series of questions to estimate the receiver’s knowledge and adjust the method of semantic information extraction according to the estimation. Due to the limited wireless resources and communication time, the size of the extracted semantic information and the number of questions that the transmitter can ask are limited. This problem is formulated as an optimization problem whose goal is to maximize worse case answer similarities over semantic generation and question selection decisions. More specifically, the transmitter aims to ask the questions to pinpoint the largest knowledge divergence and adjusts its semantic generation method to minimize this divergence. An adversarial reinforcement learning (ARL) inspired algorithm, combined with a matching network, is designed to achieve such opposite goals by searching the optimal semantic information generation scheme and question selection scheme in an adversarial manner. Simulation results demonstrate that the proposed framework can improve the semantic similarity of answers by up to 5.5% gain and can achieve up to 10.7% gain in terms of the average similarity of texts compared to the algorithm without knowledge estimation.
- Top Cited Papers
- 6 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73  (9) , 7304-7318
Compression Ratio Allocation for Probabilistic Semantic Communication With RSMA
Semantic communication is envisioned as a key technology for future wireless networks due to its high communication efficiency. However, research combining semantic communication and advanced multiple access techniques, such as rate splitting multiple access (RSMA), is still lacking. In this paper, the problem of joint communication and computation resource allocation for probabilistic semantic communication (PSCom) with RSMA is investigated. In the considered model, the base station (BS) needs to transmit a large amount of data to multiple users with 1-layer RSMA. Due to limited communication resources, the BS is required to utilize semantic communication techniques to compress the original data. In this paper, we utilize knowledge graphs to represent semantic information and employ probabilistic graphs, which are shared between the BS and users, to further compress the knowledge graphs. The BS can use the probabilistic graph to compress the data to be transmitted, while the user can recover the compressed semantic information using the same shared probabilistic graph. The additional computation power required for semantic information compression inevitably results in a reduction in transmission power due to the limited total power budget. Considering the effect of semantic compression ratio, the semantic rate expression for RSMA is first obtained. Then, based on the obtained rate expression, an optimization problem is formulated with the aim of maximizing the sum of semantic rates of all users under total power, semantic compression ratio, and rate allocation constraints. To tackle this problem, an iterative algorithm is proposed, where the semantic compression ratio subproblem is addressed using a greedy algorithm, and the rate allocation and transmit beamforming design subproblem is solved using a successive convex approximation method. Numerical results validate the effectiveness of the proposed scheme.
- 6 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Communications
- Vol. 73  (9) , 7304-7318
Top Cited PapersCompression Ratio Allocation for Probabilistic Semantic Communication With RSMA
Semantic communication is envisioned as a key technology for future wireless networks due to its high communication efficiency. However, research combining semantic communication and advanced multiple access techniques, such as rate splitting multiple access (RSMA), is still lacking. In this paper, the problem of joint communication and computation resource allocation for probabilistic semantic communication (PSCom) with RSMA is investigated. In the considered model, the base station (BS) needs to transmit a large amount of data to multiple users with 1-layer RSMA. Due to limited communication resources, the BS is required to utilize semantic communication techniques to compress the original data. In this paper, we utilize knowledge graphs to represent semantic information and employ probabilistic graphs, which are shared between the BS and users, to further compress the knowledge graphs. The BS can use the probabilistic graph to compress the data to be transmitted, while the user can recover the compressed semantic information using the same shared probabilistic graph. The additional computation power required for semantic information compression inevitably results in a reduction in transmission power due to the limited total power budget. Considering the effect of semantic compression ratio, the semantic rate expression for RSMA is first obtained. Then, based on the obtained rate expression, an optimization problem is formulated with the aim of maximizing the sum of semantic rates of all users under total power, semantic compression ratio, and rate allocation constraints. To tackle this problem, an iterative algorithm is proposed, where the semantic compression ratio subproblem is addressed using a greedy algorithm, and the rate allocation and transmit beamforming design subproblem is solved using a successive convex approximation method. Numerical results validate the effectiveness of the proposed scheme.
- 27 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 24  (8) , 6843-6857
Continual Reinforcement Learning for Digital Twin Synchronization Optimization
This article investigates the adaptive resource allocation scheme for digital twin (DT) synchronization optimization over dynamic wireless networks. In our considered model, a base station (BS) continuously collects factory physical object state data from wireless devices to build a real-time virtual DT system for factory event analysis. Due to continuous data transmission, maintaining DT synchronization must use extensive wireless resources. To address this issue, a subset of devices is selected to transmit their sensing data, and resource block (RB) allocation is optimized. This problem is formulated as a constrained Markov process (CMDP) problem that minimizes the long-term mismatch between the physical and virtual systems. To solve this CMDP, we first transform the problem into a dual problem that refines RB constraint impacts on device scheduling strategies. We then propose a continual reinforcement learning (CRL) algorithm to solve the dual problem. The CRL algorithm learns a stable policy across historical experiences for quick adaptation to dynamics in physical states and network capacity. Simulation results show that the CRL can adapt quickly to network capacity changes and reduce normalized root mean square error (NRMSE) between physical and virtual states by up to 55.2%, using the same RB number as traditional methods.
- 27 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 24  (8) , 6843-6857
Continual Reinforcement Learning for Digital Twin Synchronization Optimization
This article investigates the adaptive resource allocation scheme for digital twin (DT) synchronization optimization over dynamic wireless networks. In our considered model, a base station (BS) continuously collects factory physical object state data from wireless devices to build a real-time virtual DT system for factory event analysis. Due to continuous data transmission, maintaining DT synchronization must use extensive wireless resources. To address this issue, a subset of devices is selected to transmit their sensing data, and resource block (RB) allocation is optimized. This problem is formulated as a constrained Markov process (CMDP) problem that minimizes the long-term mismatch between the physical and virtual systems. To solve this CMDP, we first transform the problem into a dual problem that refines RB constraint impacts on device scheduling strategies. We then propose a continual reinforcement learning (CRL) algorithm to solve the dual problem. The CRL algorithm learns a stable policy across historical experiences for quick adaptation to dynamics in physical states and network capacity. Simulation results show that the CRL can adapt quickly to network capacity changes and reduce normalized root mean square error (NRMSE) between physical and virtual states by up to 55.2%, using the same RB number as traditional methods.
- 13 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24  (4) , 2894-2909
Passive Inter-Satellite Localization Accuracy Optimization in Low Earth Orbit Satellite Networks
In this paper, a passive low earth orbit (LEO) satellite localization framework is investigated. In our considered model, one active satellite and multiple passive satellites are selected to localize a target LEO satellite, where the active satellite transmits signals to the target satellite and passive satellites receive signals reflected by the target satellite. Based on the received signals, passive satellites calculate the transmission distances and send this distance information to the active satellite that will estimate the position of target satellite. Since LEO satellites are powered by the sun, the available energy that can be used for target satellite localization is limited and dynamic. Hence, the satellite selection scheme must be optimized for improving the localization accuracy under the energy consumption constraints. This problem is cast into an optimization setting with a goal of minimizing target satellite positioning error by jointly optimizing active/passive satellite selection and transmit power allocation. To solve this problem, a mixture Gaussian distribution-based reinforcement learning (MGD-RL) method is proposed. The proposed MGD-RL method enables each LEO satellite to determine whether to be an active or a passive satellite and optimize its transmit power under the energy constraints. Furthermore, the proposed MGD-RL method can approximate the probability distribution of value functions by using mixture Gaussian distributions, thus reducing the training complexity of the designed RL. Simulation results demonstrate that, compared to a value decomposition network method and independent RL method, the MGD-RL method can improve the positioning accuracy of the target LEO satellite by up to 26.8% and 48.9%.
- 13 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24  (4) , 2894-2909
Passive Inter-Satellite Localization Accuracy Optimization in Low Earth Orbit Satellite Networks
In this paper, a passive low earth orbit (LEO) satellite localization framework is investigated. In our considered model, one active satellite and multiple passive satellites are selected to localize a target LEO satellite, where the active satellite transmits signals to the target satellite and passive satellites receive signals reflected by the target satellite. Based on the received signals, passive satellites calculate the transmission distances and send this distance information to the active satellite that will estimate the position of target satellite. Since LEO satellites are powered by the sun, the available energy that can be used for target satellite localization is limited and dynamic. Hence, the satellite selection scheme must be optimized for improving the localization accuracy under the energy consumption constraints. This problem is cast into an optimization setting with a goal of minimizing target satellite positioning error by jointly optimizing active/passive satellite selection and transmit power allocation. To solve this problem, a mixture Gaussian distribution-based reinforcement learning (MGD-RL) method is proposed. The proposed MGD-RL method enables each LEO satellite to determine whether to be an active or a passive satellite and optimize its transmit power under the energy constraints. Furthermore, the proposed MGD-RL method can approximate the probability distribution of value functions by using mixture Gaussian distributions, thus reducing the training complexity of the designed RL. Simulation results demonstrate that, compared to a value decomposition network method and independent RL method, the MGD-RL method can improve the positioning accuracy of the target LEO satellite by up to 26.8% and 48.9%.
- 13 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23  (10) , 12917-12932
Jointly Optimizing Terahertz Based Sensing and Communications in Vehicular Networks: A Dynamic Graph Neural Network Approach
In this paper, the problem of vehicle service mode selection (sensing, communication, or both) and vehicle connections within terahertz (THz) enabled joint sensing and communications over vehicular networks is studied. The considered network consists of several service provider vehicles (SPVs) that can provide: 1) only sensing service, 2) only communication service, and 3) both services, sensing service request vehicles, and communication service request vehicles. Based on the vehicle network topology and their service accessibility, SPVs strategically select service request vehicles to provide sensing, communication, or both services. This problem is formulated as an optimization problem, aiming to maximize the number of successfully served vehicles by jointly determining the service mode of each SPV and its associated vehicles. To solve this problem, we propose a dynamic graph neural network (GNN) model that selects appropriate graph information aggregation functions according to the vehicle network topology, thus extracting more vehicle network information compared to traditional static GNNs that use fixed aggregation functions for different vehicle network topologies. Using the extracted vehicle network information, the service mode of each SPV and its served service request vehicles will be determined. Simulation results show that the proposed dynamic GNN based method can improve the number of successfully served vehicles by up to 17% and 28% compared to a GNN based algorithm with a fixed neural network model and a conventional optimization algorithm without using GNNs.
- 13 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23  (10) , 12917-12932
Jointly Optimizing Terahertz Based Sensing and Communications in Vehicular Networks: A Dynamic Graph Neural Network Approach
In this paper, the problem of vehicle service mode selection (sensing, communication, or both) and vehicle connections within terahertz (THz) enabled joint sensing and communications over vehicular networks is studied. The considered network consists of several service provider vehicles (SPVs) that can provide: 1) only sensing service, 2) only communication service, and 3) both services, sensing service request vehicles, and communication service request vehicles. Based on the vehicle network topology and their service accessibility, SPVs strategically select service request vehicles to provide sensing, communication, or both services. This problem is formulated as an optimization problem, aiming to maximize the number of successfully served vehicles by jointly determining the service mode of each SPV and its associated vehicles. To solve this problem, we propose a dynamic graph neural network (GNN) model that selects appropriate graph information aggregation functions according to the vehicle network topology, thus extracting more vehicle network information compared to traditional static GNNs that use fixed aggregation functions for different vehicle network topologies. Using the extracted vehicle network information, the service mode of each SPV and its served service request vehicles will be determined. Simulation results show that the proposed dynamic GNN based method can improve the number of successfully served vehicles by up to 17% and 28% compared to a GNN based algorithm with a fixed neural network model and a conventional optimization algorithm without using GNNs.
- 28 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 23  (12) , 10787-10802
Collaborative Reinforcement Learning Based Unmanned Aerial Vehicle (UAV) Trajectory Design for 3D UAV Tracking
In this paper, the problem of using one active unmanned aerial vehicle (UAV) and four passive UAVs to localize a 3D target UAV in real time is investigated. In the considered model, each passive UAV receives reflection signals from the target UAV, which are initially transmitted by the active UAV. The received reflection signals allow each passive UAV to estimate the signal transmission distance which will be transmitted to a base station (BS) for the estimation of the position of the target UAV. Due to the movement of the target UAV, each active/passive UAV must optimize its trajectory to continuously localize the target UAV. Meanwhile, since the accuracy of the distance estimation depends on the signal-to-noise ratio of the transmission signals, the active UAV must optimize its transmit power. This problem is formulated as an optimization problem whose goal is to jointly optimize the transmit power of the active UAV and trajectories of both active and passive UAVs so as to maximize the target UAV positioning accuracy. To solve this problem, a Z function decomposition based reinforcement learning (ZD-RL) method is proposed. Compared to value function decomposition based RL (VD-RL), the proposed method can find the probability distribution of the sum of future rewards to accurately estimate the expected value of the sum of future rewards thus finding better transmit power of the active UAV and trajectories for both active and passive UAVs and improving target UAV positioning accuracy. Simulation results show that the proposed ZD-RL method can reduce the positioning errors by up to 39.4% and 64.6%, compared to VD-RL and independent deep RL methods, respectively.
- 28 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 23  (12) , 10787-10802
Collaborative Reinforcement Learning Based Unmanned Aerial Vehicle (UAV) Trajectory Design for 3D UAV Tracking
In this paper, the problem of using one active unmanned aerial vehicle (UAV) and four passive UAVs to localize a 3D target UAV in real time is investigated. In the considered model, each passive UAV receives reflection signals from the target UAV, which are initially transmitted by the active UAV. The received reflection signals allow each passive UAV to estimate the signal transmission distance which will be transmitted to a base station (BS) for the estimation of the position of the target UAV. Due to the movement of the target UAV, each active/passive UAV must optimize its trajectory to continuously localize the target UAV. Meanwhile, since the accuracy of the distance estimation depends on the signal-to-noise ratio of the transmission signals, the active UAV must optimize its transmit power. This problem is formulated as an optimization problem whose goal is to jointly optimize the transmit power of the active UAV and trajectories of both active and passive UAVs so as to maximize the target UAV positioning accuracy. To solve this problem, a Z function decomposition based reinforcement learning (ZD-RL) method is proposed. Compared to value function decomposition based RL (VD-RL), the proposed method can find the probability distribution of the sum of future rewards to accurately estimate the expected value of the sum of future rewards thus finding better transmit power of the active UAV and trajectories for both active and passive UAVs and improving target UAV positioning accuracy. Simulation results show that the proposed ZD-RL method can reduce the positioning errors by up to 39.4% and 64.6%, compared to VD-RL and independent deep RL methods, respectively.