Shutao Zhang
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
Shutao Zhang has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Tsung-Hui Chang
The Chinese University of Hong Kong
3 shared publications - Qingjiang Shi
Tongji University
2 shared publications Ye Xue
Shenzhen Research Institute of Big Data, Shenzhen, China
2 shared publicationsChao Shen
Shenzhen International Center for Industrial and Applied Mathematics, Shenzhen Research Institute of Big Data, Shenzhen, China
1 shared publication- Chun Chen
China Jiliang University, College of Life Sciences, Hangzhou, 310018, China
1 shared publication Hao Wang
Shenzhen International Center for Industrial and Applied Mathematics, Shenzhen Research Institute of Big Data, Shenzhen, China
1 shared publication
Latest Publications
- 21 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24 (7) , 6069-6084
Robust Network Optimization by Deep Generative Models and Stochastic Optimization
Wireless network optimization is essential for improving the network performance in mobile communications. However, due to the stochastic nature of wireless networks, existing schemes based on analytical models and deterministic optimization are less reliable. To this end, we design a framework for robust network optimization based on deep generative models and stochastic optimization. Inspired by the powerful diffusion process, we propose a deep generative simulator to capture the statistical distribution of the network performance. By sampling from the deep generative simulator, we can alleviate the inherent uncertainty related to the network performance and devise an innovative expectation-quantile-based stochastic objective function. The inner expectation is designed for the temporal statistics, while the outer quantile is developed for the spatial statistics. This designated two-tier objective function is capable of mitigating temporal fluctuations and ensuring satisfactory network performance across most geographical grids, thereby achieving robustness. To solve this stochastic optimization problem, a smooth zeroth-order approach is introduced by taking advantage of the unique structure of quantile functions. Through theoretical performance analysis and simulation experiments with real-world datasets, we demonstrate the superiority of our approach over other baseline schemes, highlighting its practical utility in robust network optimization.
- 21 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24 (7) , 6069-6084
Robust Network Optimization by Deep Generative Models and Stochastic Optimization
Wireless network optimization is essential for improving the network performance in mobile communications. However, due to the stochastic nature of wireless networks, existing schemes based on analytical models and deterministic optimization are less reliable. To this end, we design a framework for robust network optimization based on deep generative models and stochastic optimization. Inspired by the powerful diffusion process, we propose a deep generative simulator to capture the statistical distribution of the network performance. By sampling from the deep generative simulator, we can alleviate the inherent uncertainty related to the network performance and devise an innovative expectation-quantile-based stochastic objective function. The inner expectation is designed for the temporal statistics, while the outer quantile is developed for the spatial statistics. This designated two-tier objective function is capable of mitigating temporal fluctuations and ensuring satisfactory network performance across most geographical grids, thereby achieving robustness. To solve this stochastic optimization problem, a smooth zeroth-order approach is introduced by taking advantage of the unique structure of quantile functions. Through theoretical performance analysis and simulation experiments with real-world datasets, we demonstrate the superiority of our approach over other baseline schemes, highlighting its practical utility in robust network optimization.
- 12 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24 (7) , 5508-5524
GNN-Based Structured Bayesian Inference for Multi-Grid Localized Statistical Channel Modeling
Localized statistical channel modeling (LSCM) is an efficient channel modeling framework recently proposed for wireless network optimization which learns the angular power spectrum (APS) of the downlink channel from the beam-wise reference signal receiving power (RSRP). However, the conventional LSCM is only based on RSRP measurements from one single geographical grid and ignores the inherent property of spatial consistency over wireless channels, resulting in suboptimal performance. To this end, we consider the LSCM in a manner of multiple geographical grids and further propose a novel graph-based approach for the multi-grid LSCM, called the accelerated Markovian variational Bayesian graph neural network (AMVB-GNN). The AMVB-GNN leverages a heterogeneous Markovian graph representation to capture the structured sparsity in the channel APSs and employs refined variational Bayesian inference (VBI) to learn the APSs of multiple grids. Notably, the design of AMVB-GNN eliminates the exact matrix inversion operations required in conventional VBI, thereby enhancing computational efficiency. Additionally, we demonstrate the partial permutation equivalence of AMVB-GNN, ensuring both interpretability and reliability. To address the issue of the demand for ground-truth APSs labels, we propose an unsupervised training loss function. Extensive simulation experiments validate the effectiveness and efficiency of the proposed AMVB-GNN model.
- 12 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 24 (7) , 5508-5524
GNN-Based Structured Bayesian Inference for Multi-Grid Localized Statistical Channel Modeling
Localized statistical channel modeling (LSCM) is an efficient channel modeling framework recently proposed for wireless network optimization which learns the angular power spectrum (APS) of the downlink channel from the beam-wise reference signal receiving power (RSRP). However, the conventional LSCM is only based on RSRP measurements from one single geographical grid and ignores the inherent property of spatial consistency over wireless channels, resulting in suboptimal performance. To this end, we consider the LSCM in a manner of multiple geographical grids and further propose a novel graph-based approach for the multi-grid LSCM, called the accelerated Markovian variational Bayesian graph neural network (AMVB-GNN). The AMVB-GNN leverages a heterogeneous Markovian graph representation to capture the structured sparsity in the channel APSs and employs refined variational Bayesian inference (VBI) to learn the APSs of multiple grids. Notably, the design of AMVB-GNN eliminates the exact matrix inversion operations required in conventional VBI, thereby enhancing computational efficiency. Additionally, we demonstrate the partial permutation equivalence of AMVB-GNN, ensuring both interpretability and reliability. To address the issue of the demand for ground-truth APSs labels, we propose an unsupervised training loss function. Extensive simulation experiments validate the effectiveness and efficiency of the proposed AMVB-GNN model.
- 27 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23 (6) , 5409-5424
A Physics-Based and Data-Driven Approach for Localized Statistical Channel Modeling
Localized channel modeling is crucial for offline performance optimization of wireless networks, but existing channel models are not well suited for wireless network optimization. In this paper, we propose a physics-based and data-driven localized statistical channel model for wireless network optimization. The proposed channel modeling solely relies on the reference signal receiving power (RSRP). The key is to build the statistical relationship between the RSRP and the angular power spectrum (APS). Based on it, we formulate the task of channel modeling as a sparse recovery problem where the non-zero entries of the APS indicate the channel paths’ powers and angles of departure. Although such problem typically can be handled by orthogonal matching pursuit (OMP)-type algorithms, our problem is more challenging due to the non-uniform and closely parallel columns of the coefficient matrix. To address these issues, we propose the weighted non-negative OMP (WNOMP) and the second-order-statistics-based WNOMP (SWOMP) algorithms. The WNOMP algorithm can alleviate the effect of non-uniform columns, while the SWOMP algorithm can further identify the closely parallel columns correctly. Finally, comprehensive experiments based on synthetic and real-world RSRP are presented to demonstrate that the proposed methods outperform classic methods in terms of accuracy and mean absolute error (MAE).
- 27 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23 (6) , 5409-5424
A Physics-Based and Data-Driven Approach for Localized Statistical Channel Modeling
Localized channel modeling is crucial for offline performance optimization of wireless networks, but existing channel models are not well suited for wireless network optimization. In this paper, we propose a physics-based and data-driven localized statistical channel model for wireless network optimization. The proposed channel modeling solely relies on the reference signal receiving power (RSRP). The key is to build the statistical relationship between the RSRP and the angular power spectrum (APS). Based on it, we formulate the task of channel modeling as a sparse recovery problem where the non-zero entries of the APS indicate the channel paths’ powers and angles of departure. Although such problem typically can be handled by orthogonal matching pursuit (OMP)-type algorithms, our problem is more challenging due to the non-uniform and closely parallel columns of the coefficient matrix. To address these issues, we propose the weighted non-negative OMP (WNOMP) and the second-order-statistics-based WNOMP (SWOMP) algorithms. The WNOMP algorithm can alleviate the effect of non-uniform columns, while the SWOMP algorithm can further identify the closely parallel columns correctly. Finally, comprehensive experiments based on synthetic and real-world RSRP are presented to demonstrate that the proposed methods outperform classic methods in terms of accuracy and mean absolute error (MAE).
- 1 March 2021
- Published by Elsevier in Journal of Mathematical Analysis and Applications
- Vol. 495 (2) , 124750
Extremal problems of Hardy-Littlewood-Sobolev inequalities on compact Riemannian manifolds
- 1 March 2021
- Published by Elsevier in Journal of Mathematical Analysis and Applications
- Vol. 495 (2) , 124750
Extremal problems of Hardy-Littlewood-Sobolev inequalities on compact Riemannian manifolds
- Open Access
- 3 July 2017
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 12 (7) , e0179930
Identification and validation of reference genes for qRT-PCR studies of the obligate aphid pathogenic fungus Pandora neoaphidis during different developmental stages
The selection of stable reference genes is a critical step for the accurate quantification of gene expression. To identify and validate the reference genes in Pandora neoaphidis–an obligate aphid pathogenic fungus—the expression of 13classical candidate reference genes were evaluated by quantitative real-time reverse transcriptase polymerase chain reaction(qPCR) at four developmental stages (conidia, conidia with germ tubes, short hyphae and elongated hyphae). Four statistical algorithms, including geNorm, NormFinder, BestKeeper and Delta Ct method were used to rank putative reference genes according to their expression stability and indicate the best reference gene or combination of reference genes for accurate normalization. The analysis of comprehensive ranking revealed that ACT1and 18Swas the most stably expressed genes throughout the developmental stages. To further validate the suitability of the reference genes identified in this study, the expression of cell division control protein 25 (CDC25) and Chitinase 1(CHI1) genes were used to further confirm the validated candidate reference genes. Our study presented the first systematic study of reference gene(s) selection for P. neoaphidis study and provided guidelines to obtain more accurate qPCR results for future developmental efforts.
- 3 July 2017
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 12 (7) , e0179930
Open AccessIdentification and validation of reference genes for qRT-PCR studies of the obligate aphid pathogenic fungus Pandora neoaphidis during different developmental stages
The selection of stable reference genes is a critical step for the accurate quantification of gene expression. To identify and validate the reference genes in Pandora neoaphidis–an obligate aphid pathogenic fungus—the expression of 13classical candidate reference genes were evaluated by quantitative real-time reverse transcriptase polymerase chain reaction(qPCR) at four developmental stages (conidia, conidia with germ tubes, short hyphae and elongated hyphae). Four statistical algorithms, including geNorm, NormFinder, BestKeeper and Delta Ct method were used to rank putative reference genes according to their expression stability and indicate the best reference gene or combination of reference genes for accurate normalization. The analysis of comprehensive ranking revealed that ACT1and 18Swas the most stably expressed genes throughout the developmental stages. To further validate the suitability of the reference genes identified in this study, the expression of cell division control protein 25 (CDC25) and Chitinase 1(CHI1) genes were used to further confirm the validated candidate reference genes. Our study presented the first systematic study of reference gene(s) selection for P. neoaphidis study and provided guidelines to obtain more accurate qPCR results for future developmental efforts.