Youming Sun
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Youming Sun has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
- Xiangcheng Li
Guangxi Key Laboratory of Multimedia Communications and Network Technology, Guangxi University, Nanning, China
18 shared publications - Haiqiang Chen
Guangxi University
17 shared publications - Tuanfa Qin
School of Computer and Electronic Information, Guangxi University, Nanning 530004, China
6 shared publications - Shuping Dang
Department of Electrical & Electronic Engineering, University of Bristol, Bristol BS8 1QU, UK
5 shared publications Dongri Ban
Guangxi University
4 shared publications- Tianchuan Yang
School of Electrical Engineering, Guangxi University, Nanning, China
4 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 13282-13290
Triplet-Guided Dual-Domain Manifold Reshaping for Large-Scale Incomplete Multiview Clustering
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 13282-13290
Open AccessTriplet-Guided Dual-Domain Manifold Reshaping for Large-Scale Incomplete Multiview Clustering
- Open Access
- 1 October 2026
- Published by Tsinghua University Press in Tsinghua Science and Technology
- Vol. 31 (5) , 2366-2380
On Performance of Heterogeneous Wireless-Powered Communication Systems with Co-Channel Interference
This work focuses on the study of a heterogeneous wireless-powered communication (H-WPCN) system, where a device-to-device (D2D) communication pair is overlaid with a downlink (DL) communication pair. In DL communications, the access point keeps spreading radio frequency (RF) signals to serve the DL user equipment (UE). In addition to being the data source of the UE, DL RF signals are also utilized by the D2D transmitter as the only energy source supporting D2D communications. To capture the impact of limited spectrum resource on the proposed H-WPCN system, co-channel interference is considered. Assuming the Rayleigh block fading channel for signal propagation, a tight lower bound as well as its closed-form approximation are provided for the outage probability of the DL communication link, and an exact expression is derived for the outage probability of the D2D communication link. After that, the throughput of the H-WPCN system is optimized via joint power control and time allocation, where the throughput of the D2D communication link is maximized subject to a required throughput for DL communications. Finally, our theoretical analysis is verified by numerical simulations. The simulation results show that the throughput of the H-WPCN system may not benefit from a high energy conversion efficiency.
- 1 October 2026
- Published by Tsinghua University Press in Tsinghua Science and Technology
- Vol. 31 (5) , 2366-2380
Open AccessOn Performance of Heterogeneous Wireless-Powered Communication Systems with Co-Channel Interference
This work focuses on the study of a heterogeneous wireless-powered communication (H-WPCN) system, where a device-to-device (D2D) communication pair is overlaid with a downlink (DL) communication pair. In DL communications, the access point keeps spreading radio frequency (RF) signals to serve the DL user equipment (UE). In addition to being the data source of the UE, DL RF signals are also utilized by the D2D transmitter as the only energy source supporting D2D communications. To capture the impact of limited spectrum resource on the proposed H-WPCN system, co-channel interference is considered. Assuming the Rayleigh block fading channel for signal propagation, a tight lower bound as well as its closed-form approximation are provided for the outage probability of the DL communication link, and an exact expression is derived for the outage probability of the D2D communication link. After that, the throughput of the H-WPCN system is optimized via joint power control and time allocation, where the throughput of the D2D communication link is maximized subject to a required throughput for DL communications. Finally, our theoretical analysis is verified by numerical simulations. The simulation results show that the throughput of the H-WPCN system may not benefit from a high energy conversion efficiency.
- Open Access
Task-Oriented Deep Joint Source-Channel Coding with Semantic-Aware Adaptive Quantization for Autonomous Driving
In autonomous driving and intelligent transportation, vehicles need to share visual perception information to extend sensing range, but conventional separate source-channel coding can suffer from the cliff effect under harsh vehicular channels, causing downstream perception failures. This paper proposes an end-to-end semantic communication system for autonomous-driving semantic segmentation. The system adopts SegFormer-B2 as the semantic encoder, performs feature modulation conditioned on the signal-to-noise ratio , applies semantic-aware adaptive bit-width quantization based on Gumbel-Softmax, and uses a task-oriented decoder to directly produce segmentation maps. Experiments on Cityscapes evaluate the system under additive white Gaussian noise (AWGN) and per-channel independent block Rayleigh fading channels, with comparisons against a JPEG-based separate baseline and fixed-bit-width variants. Under AWGN, the proposed system maintains a mean intersection over union above 0.70 at , while the conventional baseline fails at with an mIoU of 0.02, indicating improved robustness against the cliff effect. The results also show a non-monotonic relationship between quantization precision and segmentation performance: fixed 4-bit quantization can outperform fixed 6-bit quantization because finer quantization is more sensitive to channel noise. Under Rayleigh fading, adaptive quantization may suffer from semantic-channel mismatch, providing useful observations for future channel-aware semantic resource allocation.Open AccessTask-Oriented Deep Joint Source-Channel Coding with Semantic-Aware Adaptive Quantization for Autonomous Driving
In autonomous driving and intelligent transportation, vehicles need to share visual perception information to extend sensing range, but conventional separate source-channel coding can suffer from the cliff effect under harsh vehicular channels, causing downstream perception failures. This paper proposes an end-to-end semantic communication system for autonomous-driving semantic segmentation. The system adopts SegFormer-B2 as the semantic encoder, performs feature modulation conditioned on the signal-to-noise ratio , applies semantic-aware adaptive bit-width quantization based on Gumbel-Softmax, and uses a task-oriented decoder to directly produce segmentation maps. Experiments on Cityscapes evaluate the system under additive white Gaussian noise (AWGN) and per-channel independent block Rayleigh fading channels, with comparisons against a JPEG-based separate baseline and fixed-bit-width variants. Under AWGN, the proposed system maintains a mean intersection over union above 0.70 at , while the conventional baseline fails at with an mIoU of 0.02, indicating improved robustness against the cliff effect. The results also show a non-monotonic relationship between quantization precision and segmentation performance: fixed 4-bit quantization can outperform fixed 6-bit quantization because finer quantization is more sensitive to channel noise. Under Rayleigh fading, adaptive quantization may suffer from semantic-channel mismatch, providing useful observations for future channel-aware semantic resource allocation. - 2 July 2026
- Published by Elsevier in Information Fusion
- Vol. 137
Joint tensor self-representation and discriminative feature extraction for multi-view clustering
- 2 July 2026
- Published by Elsevier in Information Fusion
- Vol. 137
Joint tensor self-representation and discriminative feature extraction for multi-view clustering
- 22 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. PP (10518215) , 1
Density Peak Anchor Selection for Multi-View Clustering
Multi-view clustering (MVC) typically enhances clustering performance by leveraging consistent and complementary information across views. With increasing data scale, many MVC methods employ few anchors to represent data, effectively reducing complexity. Traditional anchor selection methods include random sampling and k-means. Recent methods leverage variance or inter-class error to guide anchor selection. However, their anchors are selected sequentially, where each newly selected anchor focuses on divergences from the previous anchor. This local perspective leads to unstable anchor coverage and fails to fully represent complex cluster shapes. An anchor selection algorithm, termed DPAS, is proposed to solve this problem. From a novel density perspective, it selects density peaks satisfying criteria as anchors and reduces complexity based on granular balls. As a result, the efficient DPAS exhibits stronger representativeness and robustness. Furthermore, the order and number of anchors selected from different views typically need to be aligned. The assumptions of existing alignment methods are often too restrictive, compromising the quality of alignment. Thus, an alignment method FAA based on flexible assumptions is designed to improve generalization. In this paper, we propose a new fast MVC method without alternating optimization named DPAS-MC. First, DPAS is applied to select high-quality anchors for views. Then, anchor graphs are computed, aligned, and fused via FAA. Finally, a low-rank tensor is calculated to obtain clustering results. Experiments show that our methods outperform existing state-of-the-art methods in performance and efficiency. The code is available at https://github.com/ytccyw.
- 22 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. PP (10518215) , 1
Density Peak Anchor Selection for Multi-View Clustering
Multi-view clustering (MVC) typically enhances clustering performance by leveraging consistent and complementary information across views. With increasing data scale, many MVC methods employ few anchors to represent data, effectively reducing complexity. Traditional anchor selection methods include random sampling and k-means. Recent methods leverage variance or inter-class error to guide anchor selection. However, their anchors are selected sequentially, where each newly selected anchor focuses on divergences from the previous anchor. This local perspective leads to unstable anchor coverage and fails to fully represent complex cluster shapes. An anchor selection algorithm, termed DPAS, is proposed to solve this problem. From a novel density perspective, it selects density peaks satisfying criteria as anchors and reduces complexity based on granular balls. As a result, the efficient DPAS exhibits stronger representativeness and robustness. Furthermore, the order and number of anchors selected from different views typically need to be aligned. The assumptions of existing alignment methods are often too restrictive, compromising the quality of alignment. Thus, an alignment method FAA based on flexible assumptions is designed to improve generalization. In this paper, we propose a new fast MVC method without alternating optimization named DPAS-MC. First, DPAS is applied to select high-quality anchors for views. Then, anchor graphs are computed, aligned, and fused via FAA. Finally, a low-rank tensor is calculated to obtain clustering results. Experiments show that our methods outperform existing state-of-the-art methods in performance and efficiency. The code is available at https://github.com/ytccyw.
- 27 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 30 (10897798) , 1920-1924
Graph Attention Network-Aided Message-Passing Neural Decoder for Polar Codes Based on Sparse-Pruned Tanner Graphs
In this letter, we propose a fully trainable graph neural network (GNN) decoder for short polar codes, underpinned by a novel graph attention network (GAT)-aided message-passing (GAT-MP) framework. This framework aligns message passing neural network (MPNN) with the belief propagation (BP) algorithm, leveraging graph attention mechanism to enhance the decoding process. To facilitate effective learning, we introduce a tailored two-stage train-then-fine-tune (TTF) methodology, along-side a specialized training algorithm. Simulation results show that the proposed method significantly outperforms advanced BP decoders while offering competitive complexity. Furthermore, it matches or exceeds the performance of successive-cancellation list (SCL) and state-of-the-art (SOTA) Transformer-based methods, with a significantly smaller model size than the latter.
- 27 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 30 (10897798) , 1920-1924
Graph Attention Network-Aided Message-Passing Neural Decoder for Polar Codes Based on Sparse-Pruned Tanner Graphs
In this letter, we propose a fully trainable graph neural network (GNN) decoder for short polar codes, underpinned by a novel graph attention network (GAT)-aided message-passing (GAT-MP) framework. This framework aligns message passing neural network (MPNN) with the belief propagation (BP) algorithm, leveraging graph attention mechanism to enhance the decoding process. To facilitate effective learning, we introduce a tailored two-stage train-then-fine-tune (TTF) methodology, along-side a specialized training algorithm. Simulation results show that the proposed method significantly outperforms advanced BP decoders while offering competitive complexity. Furthermore, it matches or exceeds the performance of successive-cancellation list (SCL) and state-of-the-art (SOTA) Transformer-based methods, with a significantly smaller model size than the latter.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 5522-5526
UC-STARJSCC: Unified Conditional Adaptation for Cross-Channel Image Semantic Transmission
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 5522-5526
UC-STARJSCC: Unified Conditional Adaptation for Cross-Channel Image Semantic Transmission
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 4230-4234
Lightweight and Channel-Adaptive Deep Joint Source-Channel Coding for Wireless Image Transmission
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 4230-4234
Lightweight and Channel-Adaptive Deep Joint Source-Channel Coding for Wireless Image Transmission
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 1
Adversarial Robustness of Channel-Matched DeepJSCC over MIMO Fading
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 1
Adversarial Robustness of Channel-Matched DeepJSCC over MIMO Fading