Huei-Wen Ferng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Huei-Wen Ferng has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
Alexander Leonovich
Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan
4 shared publications- Amin Lotfolahi
Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan
3 shared publications - Jeng-Ji Huang
National Taiwan Normal University, Taipei, Taiwan
3 shared publications - Mingfu Li
Department of Electrical Engineering, School of Electrical and Computer Engineering, College of Engineering, Chang Gung University, Guishan District, Taoyuan 33302, Taiwan
3 shared publications - Binayak Kar
National Taiwan University of Science and Technology
2 shared publications Cheng-Ching Peng
The Department of Computer Science and Information Engineering , National Taiwan University of Science and Technology , Taipei , Taiwan, R.O.C.
2 shared publications
Latest Publications
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (18) , 43319-43332
PTFTL: Personalized Transformer-Based Federated Transfer Learning With Projection Layer for Malware Detection in Edge--IoT Networks
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (18) , 43319-43332
PTFTL: Personalized Transformer-Based Federated Transfer Learning With Projection Layer for Malware Detection in Edge--IoT Networks
- Open Access
- 21 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 149260-149272
Low-Complexity Energy Antenna Allocation for Wireless-Powered IoT Systems
To improve the performance of a wireless-powered Internet of Things (IoT) system, a multi-source wireless power transfer technique based on the multi-antenna beamforming architecture has emerged as an attractive solution. However, the distance or phase diversity among power beams from various antennas to a wireless device may degrade the energy transfer efficiency. Thus, this paper considers the energy antenna resource allocation problem in a wireless-powered IoT system with multiple wireless devices and multi-antenna energy transmitters. To maximize the system objective function such as the energy transfer efficiency, spectral efficiency, or energy spectral efficiency, we designed a greedy-based antenna allocation algorithm with polynomial time complexity by taking the distance diversity and phase differences among various energy antennas into account. According to numerous experimental tests, the proposed greedy with fine-tuning algorithm achieves a near-optimal solution while preserving very low computation complexity relative to the backtracking and dynamic programming-based algorithms. Our design makes the real-time energy antenna resource allocation in a wireless-powered IoT system feasible.
- 21 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 149260-149272
Open AccessLow-Complexity Energy Antenna Allocation for Wireless-Powered IoT Systems
To improve the performance of a wireless-powered Internet of Things (IoT) system, a multi-source wireless power transfer technique based on the multi-antenna beamforming architecture has emerged as an attractive solution. However, the distance or phase diversity among power beams from various antennas to a wireless device may degrade the energy transfer efficiency. Thus, this paper considers the energy antenna resource allocation problem in a wireless-powered IoT system with multiple wireless devices and multi-antenna energy transmitters. To maximize the system objective function such as the energy transfer efficiency, spectral efficiency, or energy spectral efficiency, we designed a greedy-based antenna allocation algorithm with polynomial time complexity by taking the distance diversity and phase differences among various energy antennas into account. According to numerous experimental tests, the proposed greedy with fine-tuning algorithm achieves a near-optimal solution while preserving very low computation complexity relative to the backtracking and dynamic programming-based algorithms. Our design makes the real-time energy antenna resource allocation in a wireless-powered IoT system feasible.
- 1 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 13, 469-484
DRL-Based Resource Allocation in NOMA-Aided Industrial IoT Towards Energy Productivity Maximization
This paper addresses the joint optimization problem of computing and communication resource allocation in a non-orthogonal multiple access (NOMA) supported industrial Internet of things (IoT) system. We introduce a decentralized deep reinforcement learning (DRL) based mechanism that collaboratively maximizes energy productivity (EP). Specifically, optimizing EP involves maximizing the amount of successfully processed bits while simultaneously minimizing task processing delay, energy consumption, and task drop ratio. With this goal, we design a comprehensive model that incorporates multiple factors, including NOMA-specific constraints, task queueing, partial processing, and multitasking, where each device can handle multiple concurrent tasks. Next, we propose a dual-decentralized multi-agent proximal policy optimization (MAPPO) based algorithm, where one MAPPO algorithm focuses explicitly on computing resource allocation for task processing at the edge servers, while the other MAPPO algorithm precisely manages communication resource allocation. Additionally, each MAPPO algorithm is equipped with a small residual network (ResNet) and a recurrent neural network (RNN) to effectively capture spatial features from complex channel conditions and the evolving task flow. Through extensive simulations, we validate the effectiveness of our resultant mechanism under various environmental conditions and demonstrate its superiority over the closely related mechanisms in the literature.
- 1 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 13, 469-484
DRL-Based Resource Allocation in NOMA-Aided Industrial IoT Towards Energy Productivity Maximization
This paper addresses the joint optimization problem of computing and communication resource allocation in a non-orthogonal multiple access (NOMA) supported industrial Internet of things (IoT) system. We introduce a decentralized deep reinforcement learning (DRL) based mechanism that collaboratively maximizes energy productivity (EP). Specifically, optimizing EP involves maximizing the amount of successfully processed bits while simultaneously minimizing task processing delay, energy consumption, and task drop ratio. With this goal, we design a comprehensive model that incorporates multiple factors, including NOMA-specific constraints, task queueing, partial processing, and multitasking, where each device can handle multiple concurrent tasks. Next, we propose a dual-decentralized multi-agent proximal policy optimization (MAPPO) based algorithm, where one MAPPO algorithm focuses explicitly on computing resource allocation for task processing at the edge servers, while the other MAPPO algorithm precisely manages communication resource allocation. Additionally, each MAPPO algorithm is equipped with a small residual network (ResNet) and a recurrent neural network (RNN) to effectively capture spatial features from complex channel conditions and the evolving task flow. Through extensive simulations, we validate the effectiveness of our resultant mechanism under various environmental conditions and demonstrate its superiority over the closely related mechanisms in the literature.
- 20 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Green Communications and Networking
- Vol. 8 (1) , 265-278
A Multi-Agent Proximal Policy Optimized Joint Mechanism in mmWave HetNets With CoMP Toward Energy Efficiency Maximization
A novel cluster-based traffic offloading and user association (UA) algorithm alongside a multi-agent deep reinforcement learning (DRL) based base station (BS) activation mechanism is proposed in this paper. Our design aims to maximize the energy efficiency (EE) of the heterogeneous network (HetNet) while maintaining high quality of service (QoS). By taking advantage of the dense deployment of BSs in a HetNet, a clustering algorithm is first proposed to facilitate traffic offloading among BSs. Then, a multi-agent proximal policy optimization (MAPPO) based DRL algorithm is employed to trigger the BS activation decision based on the current environmental condition. Finally, a UA algorithm is deployed to improve further the (normalized) data rate of all users, known as the (normalized) sum rate. Via simulation, we show that our proposed mechanism can remarkably enhance the EE and excel over the closely related mechanisms. It satisfies the required data rate, improves the sum rate, and exhibits excellent scalability when many BSs are deployed.
- 20 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Green Communications and Networking
- Vol. 8 (1) , 265-278
A Multi-Agent Proximal Policy Optimized Joint Mechanism in mmWave HetNets With CoMP Toward Energy Efficiency Maximization
A novel cluster-based traffic offloading and user association (UA) algorithm alongside a multi-agent deep reinforcement learning (DRL) based base station (BS) activation mechanism is proposed in this paper. Our design aims to maximize the energy efficiency (EE) of the heterogeneous network (HetNet) while maintaining high quality of service (QoS). By taking advantage of the dense deployment of BSs in a HetNet, a clustering algorithm is first proposed to facilitate traffic offloading among BSs. Then, a multi-agent proximal policy optimization (MAPPO) based DRL algorithm is employed to trigger the BS activation decision based on the current environmental condition. Finally, a UA algorithm is deployed to improve further the (normalized) data rate of all users, known as the (normalized) sum rate. Via simulation, we show that our proposed mechanism can remarkably enhance the EE and excel over the closely related mechanisms. It satisfies the required data rate, improves the sum rate, and exhibits excellent scalability when many BSs are deployed.
- 27 September 2023
- Published by Elsevier in Vehicular Communications
- Vol. 44
Security frameworks without the third-party infrastructure in VANETs with detailed security strength and overhead analyses
- 27 September 2023
- Published by Elsevier in Vehicular Communications
- Vol. 44
Security frameworks without the third-party infrastructure in VANETs with detailed security strength and overhead analyses
- 11 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 22 (22) , 22255-22263
Routing Protocol for a Heterogeneous MSN With an Intermittent Mobile Sink
Choosing the first broadcaster to be the parent node as proposed by the leader-based routing (LBR) enables its Polling packet to be received earliest in a heterogeneous mobile sensor network (MSN) with short and intermittent availability of the mobile sink but may lead to an unreachable parent node. Furthermore, employing the three-way-handshaking as proposed by the link quality indicator (LQI)-based beaconless routing (LQI-BLR) may lead to a longer packet delivery delay. To avoid the aforementioned drawbacks simultaneously, the transmission power of the candidate parent node is included in the Polling packet in our proposed protocol. Along with the LQI, a mobile sensor can estimate its distance to the candidate parent node and its packet-receive ratio (PRR). If the PRR is higher than a preset threshold, the candidate can be chosen accordingly. Via simulations, we demonstrate that our proposed protocol can outperform LBR and LQI-BLR in a heterogeneous environment. In terms of the packet delivery ratio, our proposed protocol is able to achieve at least 35% and 20% of improvement when compared to LBR and LQI-BLR, respectively. Regarding the energy consumption and packet delivery delay, more than 20% improvement can be observed.
- 11 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 22 (22) , 22255-22263
Routing Protocol for a Heterogeneous MSN With an Intermittent Mobile Sink
Choosing the first broadcaster to be the parent node as proposed by the leader-based routing (LBR) enables its Polling packet to be received earliest in a heterogeneous mobile sensor network (MSN) with short and intermittent availability of the mobile sink but may lead to an unreachable parent node. Furthermore, employing the three-way-handshaking as proposed by the link quality indicator (LQI)-based beaconless routing (LQI-BLR) may lead to a longer packet delivery delay. To avoid the aforementioned drawbacks simultaneously, the transmission power of the candidate parent node is included in the Polling packet in our proposed protocol. Along with the LQI, a mobile sensor can estimate its distance to the candidate parent node and its packet-receive ratio (PRR). If the PRR is higher than a preset threshold, the candidate can be chosen accordingly. Via simulations, we demonstrate that our proposed protocol can outperform LBR and LQI-BLR in a heterogeneous environment. In terms of the packet delivery ratio, our proposed protocol is able to achieve at least 35% and 20% of improvement when compared to LBR and LQI-BLR, respectively. Regarding the energy consumption and packet delivery delay, more than 20% improvement can be observed.
- Open Access
On-Demand Energy Transfer and Energy-Aware Polling-Based MAC for Wireless Powered Sensor Networks
To improve the performance of the wireless powered sensor network (WPSN), this paper proposes a frequency division duplex (FDD)-based on-demand energy transfer protocol and an energy-aware polling-based medium access control (MAC) protocol, called composite energy and data first (CEDF), by using the numbers of data packets and energy packets to determine polling priorities. The performance of the proposed MAC protocol, i.e., CEDF, along with the on-demand energy transfer protocol was evaluated through simulations, with comparison to the closely related protocols such as the round robin (RR) and data first (DF) polling protocols. Compared with RR and DF, our proposed CEDF performs much better in terms of throughput, data packet loss rate, and delay. Additionally, the doubly near–far problem in WPSNs under our proposed on-demand energy transfer protocol and CEDF was investigated to come up with good solutions to alleviate such a problem.
Open AccessOn-Demand Energy Transfer and Energy-Aware Polling-Based MAC for Wireless Powered Sensor Networks
To improve the performance of the wireless powered sensor network (WPSN), this paper proposes a frequency division duplex (FDD)-based on-demand energy transfer protocol and an energy-aware polling-based medium access control (MAC) protocol, called composite energy and data first (CEDF), by using the numbers of data packets and energy packets to determine polling priorities. The performance of the proposed MAC protocol, i.e., CEDF, along with the on-demand energy transfer protocol was evaluated through simulations, with comparison to the closely related protocols such as the round robin (RR) and data first (DF) polling protocols. Compared with RR and DF, our proposed CEDF performs much better in terms of throughput, data packet loss rate, and delay. Additionally, the doubly near–far problem in WPSNs under our proposed on-demand energy transfer protocol and CEDF was investigated to come up with good solutions to alleviate such a problem.
- 15 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 70 (12) , 13270-13280
QoS Violation Probability Minimization in Federating Vehicular-Fogs With Cloud and Edge Systems
Due to the fact that most user equipments (UEs) do not have enough computation power, time-sensitive tasks should be offloaded to other computation resources. In such cases, a cloud is an ideal destination to which computation tasks can be offloaded because of its huge data processing power. However, the distance between a UE and the cloud may increase the latency of data transmission. At the same time, edges and vehicular-fogs (consisting of electric vehicles with computing resources) are closer to UEs and mostly remain under-utilized. To address this issue, a federated architecture with UEs, vehicular-fogs (parking lot fog and traffic intersection fog), edges, and cloud is proposed along with a probabilistic offloading strategy. The quality of service (QoS) violation probability can be minimized by finding the optimal offloading probabilities while satisfying the delay constraint through an optimal offloading probability estimation algorithm based on the subgradient method proposed in this paper. Our results show that the QoS violation probability and the average waiting time can be decreased by 10%-12% and 45%, respectively, for a federated architecture with two vehicular-fogs as compared to an architecture without any fogs. As offloading to edges and cloud has longer communication delays than offloading to vehicular-fogs, the offloading probabilities to the edges in the architecture with two vehicular-fogs in the case of heavy traffic are reduced by nearly 35% as compared to the architecture without any fogs.
- 15 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 70 (12) , 13270-13280
QoS Violation Probability Minimization in Federating Vehicular-Fogs With Cloud and Edge Systems
Due to the fact that most user equipments (UEs) do not have enough computation power, time-sensitive tasks should be offloaded to other computation resources. In such cases, a cloud is an ideal destination to which computation tasks can be offloaded because of its huge data processing power. However, the distance between a UE and the cloud may increase the latency of data transmission. At the same time, edges and vehicular-fogs (consisting of electric vehicles with computing resources) are closer to UEs and mostly remain under-utilized. To address this issue, a federated architecture with UEs, vehicular-fogs (parking lot fog and traffic intersection fog), edges, and cloud is proposed along with a probabilistic offloading strategy. The quality of service (QoS) violation probability can be minimized by finding the optimal offloading probabilities while satisfying the delay constraint through an optimal offloading probability estimation algorithm based on the subgradient method proposed in this paper. Our results show that the QoS violation probability and the average waiting time can be decreased by 10%-12% and 45%, respectively, for a federated architecture with two vehicular-fogs as compared to an architecture without any fogs. As offloading to edges and cloud has longer communication delays than offloading to vehicular-fogs, the offloading probabilities to the edges in the architecture with two vehicular-fogs in the case of heavy traffic are reduced by nearly 35% as compared to the architecture without any fogs.
- 5 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 70 (10) , 9741-9751
An Improved Traffic Rerouting Strategy Using Real-Time Traffic Information and Decisive Weights
To alleviate possible traffic congestion during rush hours, an improved traffic rerouting strategy designed for (or interplayed with) the vehicular ad hoc networks (VANETs) is proposed in this paper. The proposed strategy considers not only the real-time traffic information but also the maximization of road utilization to reach an optimized solution. Furthermore, an alternative path selection is incorporated so that an efficient rerouting path is determined according to the destination of the vehicle, the density and the capacity of the road, and the speed. Via simulations, we can successfully show that our proposed strategy significantly outperforms the closely related strategies in the literature in terms of the travel time, CO $_{2}$ emission, and fuel consumption, in particular, when heavy traffic is involved. Specifically, our proposed strategy can gain up to 20% of reduction in the average travel time in a realistic scenario to be observed later. Additionally, our proposed strategy gains 19% at least of improvement in the case with different road networks and 28.6% at least of improvement in the case with different percentages of reduction of traffic lights as compared to the closely related strategies as shown in the Appendices given in the supplementary downloadable material.
- 5 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 70 (10) , 9741-9751
An Improved Traffic Rerouting Strategy Using Real-Time Traffic Information and Decisive Weights
To alleviate possible traffic congestion during rush hours, an improved traffic rerouting strategy designed for (or interplayed with) the vehicular ad hoc networks (VANETs) is proposed in this paper. The proposed strategy considers not only the real-time traffic information but also the maximization of road utilization to reach an optimized solution. Furthermore, an alternative path selection is incorporated so that an efficient rerouting path is determined according to the destination of the vehicle, the density and the capacity of the road, and the speed. Via simulations, we can successfully show that our proposed strategy significantly outperforms the closely related strategies in the literature in terms of the travel time, CO $_{2}$ emission, and fuel consumption, in particular, when heavy traffic is involved. Specifically, our proposed strategy can gain up to 20% of reduction in the average travel time in a realistic scenario to be observed later. Additionally, our proposed strategy gains 19% at least of improvement in the case with different road networks and 28.6% at least of improvement in the case with different percentages of reduction of traffic lights as compared to the closely related strategies as shown in the Appendices given in the supplementary downloadable material.
- 17 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. PP (15361233) , 1
Beaconless Geographical Routing Protocol for a Heterogeneous MSN
Traditional beaconless geographical routing (BGR) protocols perform poorly in a heterogeneous environment due to the consideration of location solely with an assumption of homogeneous transmission. Therefore, we propose a novel BGR with consideration of the locations of candidate forwarders and different transmission ranges. Our proposed protocol, namely, the heterogeneity-aware BGR (HA-BGR), is designed for a heterogeneous mobile sensor network (MSN), where the transmission ranges of mobile sensors are different. In HA-BGR, a node takes its distance, transmission range, and progress to the destination into consideration before setting itself to be a candidate next hop and remains active. Examined by analytical and extensive simulation results, HA-BGR not only needs fewer next-hop candidates but also achieves better performance in the end-to-end delivery ratio and latency as compared to the recently proposed convex-lense decision region BGR (CDR-BGR).
- 17 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. PP (15361233) , 1
Beaconless Geographical Routing Protocol for a Heterogeneous MSN
Traditional beaconless geographical routing (BGR) protocols perform poorly in a heterogeneous environment due to the consideration of location solely with an assumption of homogeneous transmission. Therefore, we propose a novel BGR with consideration of the locations of candidate forwarders and different transmission ranges. Our proposed protocol, namely, the heterogeneity-aware BGR (HA-BGR), is designed for a heterogeneous mobile sensor network (MSN), where the transmission ranges of mobile sensors are different. In HA-BGR, a node takes its distance, transmission range, and progress to the destination into consideration before setting itself to be a candidate next hop and remains active. Examined by analytical and extensive simulation results, HA-BGR not only needs fewer next-hop candidates but also achieves better performance in the end-to-end delivery ratio and latency as compared to the recently proposed convex-lense decision region BGR (CDR-BGR).