Shucheng Li
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
Shucheng Li has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Hao Wu
Nanjing University
10 shared publications - Feng Lyu
Central South University
8 shared publications - Fengyuan Xu
National Key Lab for Novel Software Technology, Nanjing University, Nanjing, Jiangsu, China
6 shared publications - Sheng Zhong
State Key Laboratory of Novel Software Technology, Nanjing University, Nanjing, Jiangsu, China
5 shared publications - Fan Wu
Central South University, Changsha, China
4 shared publications - Boyu Chang
National Key Lab for Novel Software Technology, Nanjing University, Nanjing, China
2 shared publications
Latest Publications
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (12) , 24073-24087
Scaling LLMs via Collaboration: Efficient LLM Inference Across Resource-Constrained Devices
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (12) , 24073-24087
Scaling LLMs via Collaboration: Efficient LLM Inference Across Resource-Constrained Devices
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (11) , 20056-20073
FedPKD + : A Prototype-Based Framework for Efficient and Flexible Heterogeneous Federated Learning
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (11) , 20056-20073
FedPKD + : A Prototype-Based Framework for Efficient and Flexible Heterogeneous Federated Learning
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1994-2004
SynDiSC: High-Quality Tabular Data Synthesis with Distributional and Semantic Consistency
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1994-2004
Open AccessSynDiSC: High-Quality Tabular Data Synthesis with Distributional and Semantic Consistency
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1005-1015
Seeing the Whole Through the Parts: Discovering Objects through Semantic Part Mining in Weak Supervision
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1005-1015
Open AccessSeeing the Whole Through the Parts: Discovering Objects through Semantic Part Mining in Weak Supervision
- Top Cited Papers
- 12 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 3460-3475
U-Mesh + : Terrain-Aware, Robust, and Cost-Efficient UAV-Mesh Network Deployment for Inspection Tasks in Remote Areas
Powerline inspection with UAVs significantly improves efficiency and safety in remote areas. However, the lack of cellular infrastructure necessitates the use of UAV-mesh networks, whose deployment presents challenges in jointly optimizing coverage, node load, robustness, and cost under complex terrain constraints. In this paper, we investigate the computational complexity of this deployment problem by formulating it as a multi-objective optimization task and proving its NP-hardness. To address this, we present U-Mesh+, a terrain-aware, robust, and cost-efficient deployment framework that integrates four key components: (i) identifying line-of-sight and non-line-of-sight links to model terrain-induced communication constraints; (ii) NetCons for cost-effective coverage and connectivity network topology construction; (iii) NetOpt for network resilience and balance node-level load improvement without extra cost; and (iv) NetEnh for service availability enhancement via targeted local refinements. We implement U-Mesh+ in a real-world 270km2 mountainous forest with 174 power towers and 48km of transmission lines. Extensive experiments demonstrate its efficacy in terms of deployment cost and network performance. On-site network data from the deployed wireless network further validate its effectiveness and scalability under real-world conditions.
- 12 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 3460-3475
Top Cited PapersU-Mesh + : Terrain-Aware, Robust, and Cost-Efficient UAV-Mesh Network Deployment for Inspection Tasks in Remote Areas
Powerline inspection with UAVs significantly improves efficiency and safety in remote areas. However, the lack of cellular infrastructure necessitates the use of UAV-mesh networks, whose deployment presents challenges in jointly optimizing coverage, node load, robustness, and cost under complex terrain constraints. In this paper, we investigate the computational complexity of this deployment problem by formulating it as a multi-objective optimization task and proving its NP-hardness. To address this, we present U-Mesh+, a terrain-aware, robust, and cost-efficient deployment framework that integrates four key components: (i) identifying line-of-sight and non-line-of-sight links to model terrain-induced communication constraints; (ii) NetCons for cost-effective coverage and connectivity network topology construction; (iii) NetOpt for network resilience and balance node-level load improvement without extra cost; and (iv) NetEnh for service availability enhancement via targeted local refinements. We implement U-Mesh+ in a real-world 270km2 mountainous forest with 174 power towers and 48km of transmission lines. Extensive experiments demonstrate its efficacy in terms of deployment cost and network performance. On-site network data from the deployed wireless network further validate its effectiveness and scalability under real-world conditions.
- 18 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 2332-2346
Game in Motion: Heterogeneous Task Offloading in Dynamic Vehicular Edge Computing
The rise of vehicular edge computing (VEC) enables vehicles to offload resource-intensive tasks to roadside units (RSUs), improving efficiency and reducing latency. However, high mobility, dynamic resource availability, and heterogeneous quality-of-experience (QoE) requirements make task offloading and coordination highly challenging. In this paper, we investigate the heterogeneous task offloading problem in dynamic VEC environments by proposing a two-stage optimization framework, TOVEC. Our TOVEC decouples the spatio-temporally coupled decision space into two tractable subproblems and solves it with a two-stage design. In the first stage, we employ a TD3-based deep reinforcement learning algorithm to handle RSU-channel access decisions under dynamic network conditions. In the second stage, we formulate the interaction between vehicular users (VUs) and RSUs as a Stackelberg game, enabling joint task scheduling and dynamic pricing that balances VUs’ QoE optimization and RSUs’ revenue maximization. We theoretically prove the existence of a Stackelberg equilibrium and validate our TOVEC using real-world vehicular traces. The experimental results show that our method has improved the QoE (measured by task delay and energy cost) of VU and the benefits of RSU by 6.8%-41.3% and 2.2%-117.4%, respectively, compared with the baseline schemes.
- 18 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 2332-2346
Game in Motion: Heterogeneous Task Offloading in Dynamic Vehicular Edge Computing
The rise of vehicular edge computing (VEC) enables vehicles to offload resource-intensive tasks to roadside units (RSUs), improving efficiency and reducing latency. However, high mobility, dynamic resource availability, and heterogeneous quality-of-experience (QoE) requirements make task offloading and coordination highly challenging. In this paper, we investigate the heterogeneous task offloading problem in dynamic VEC environments by proposing a two-stage optimization framework, TOVEC. Our TOVEC decouples the spatio-temporally coupled decision space into two tractable subproblems and solves it with a two-stage design. In the first stage, we employ a TD3-based deep reinforcement learning algorithm to handle RSU-channel access decisions under dynamic network conditions. In the second stage, we formulate the interaction between vehicular users (VUs) and RSUs as a Stackelberg game, enabling joint task scheduling and dynamic pricing that balances VUs’ QoE optimization and RSUs’ revenue maximization. We theoretically prove the existence of a Stackelberg equilibrium and validate our TOVEC using real-world vehicular traces. The experimental results show that our method has improved the QoE (measured by task delay and energy cost) of VU and the benefits of RSU by 6.8%-41.3% and 2.2%-117.4%, respectively, compared with the baseline schemes.
- Open Access
- 3 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1201-1203
Demo: Task Cooperation for Urban Unmanned Sanitation Vehicles
- 3 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1201-1203
Open AccessDemo: Task Cooperation for Urban Unmanned Sanitation Vehicles
- 3 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (4) , 5393-5405
H2O: Heterogeneity-Aware Hierarchical Orchestration for Memory-Efficient On-Device LLM Inference
On-device Large Language Model (LLM) inference enables private, personalized AI but faces memory constraints. Despite memory optimization efforts, scaling laws continue to increase model sizes and memory pressure. In this paper, we revisit the core memory bottlenecks in on-device LLM inference and conduct a comprehensive analysis of mainstream optimization techniques. We uncover several overlooked inefficiencies: (1) model weights, not KV caches, dominate memory usage; (2) weight sparsity remains underutilized; (3) OS-level memory behaviors cause redundancy; and (4) naive weight loading leads to excessive memory residency. To address these challenges, we propose H2O, a heterogeneity-aware hierarchical orchestration framework for memory-efficient on-device LLM inference. H2O introduces three key techniques, including hierarchical weight orchestration to reduce redundant memory retention, zero-copy I/O–compute parallelism for safe and efficient memory reuse, and heterogeneity-aware inference planning to adapt to diverse mobile hardware constraints. Extensive experimental results show that H2O reduces peak memory usage by up to 60%, eliminates out-of-memory (OOM) failures for 7 B–13 B models, and improves inference latency by 34%–94% under tight memory budgets.
- 3 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (4) , 5393-5405
H2O: Heterogeneity-Aware Hierarchical Orchestration for Memory-Efficient On-Device LLM Inference
On-device Large Language Model (LLM) inference enables private, personalized AI but faces memory constraints. Despite memory optimization efforts, scaling laws continue to increase model sizes and memory pressure. In this paper, we revisit the core memory bottlenecks in on-device LLM inference and conduct a comprehensive analysis of mainstream optimization techniques. We uncover several overlooked inefficiencies: (1) model weights, not KV caches, dominate memory usage; (2) weight sparsity remains underutilized; (3) OS-level memory behaviors cause redundancy; and (4) naive weight loading leads to excessive memory residency. To address these challenges, we propose H2O, a heterogeneity-aware hierarchical orchestration framework for memory-efficient on-device LLM inference. H2O introduces three key techniques, including hierarchical weight orchestration to reduce redundant memory retention, zero-copy I/O–compute parallelism for safe and efficient memory reuse, and heterogeneity-aware inference planning to adapt to diverse mobile hardware constraints. Extensive experimental results show that H2O reduces peak memory usage by up to 60%, eliminates out-of-memory (OOM) failures for 7 B–13 B models, and improves inference latency by 34%–94% under tight memory budgets.
- Open Access
- 30 October 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Knowledge Discovery From Data
- Vol. 18 (9) , 1-21
Multi-Label and Evolvable Dataset Preparation for Web-Based Object Detection
In this article, we focus on the emerging field of web-based object detection, which has gained considerable attention due to its ability to utilize large amounts of web data for training, thus eliminating the need for labor-intensive manual annotations. However, the noisy and ever-evolving nature of web data poses challenges in preparing high-quality datasets for web-based object detection. To address these challenges, we propose a fully automatic dataset preparation method in this article. Our proposed method incorporates a hierarchical clustering module that assigns multiple precise labels to each image. This module is based on our observation that web image data exhibits different distributions at varying granularities. Furthermore, an evolutionary relabeling module ensures the adaptability of both the prepared dataset and trained detection models to the ever-evolving web data. Extensive experiments demonstrate that our method outperforms other web-based methods, and achieves a comparable performance to those manually labeled benchmark datasets.
- 30 October 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Knowledge Discovery From Data
- Vol. 18 (9) , 1-21
Open AccessMulti-Label and Evolvable Dataset Preparation for Web-Based Object Detection
In this article, we focus on the emerging field of web-based object detection, which has gained considerable attention due to its ability to utilize large amounts of web data for training, thus eliminating the need for labor-intensive manual annotations. However, the noisy and ever-evolving nature of web data poses challenges in preparing high-quality datasets for web-based object detection. To address these challenges, we propose a fully automatic dataset preparation method in this article. Our proposed method incorporates a hierarchical clustering module that assigns multiple precise labels to each image. This module is based on our observation that web image data exhibits different distributions at varying granularities. Furthermore, an evolutionary relabeling module ensures the adaptability of both the prepared dataset and trained detection models to the ever-evolving web data. Extensive experiments demonstrate that our method outperforms other web-based methods, and achieves a comparable performance to those manually labeled benchmark datasets.
- Open Access
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1839-1847
CoAst: Validation-Free Contribution Assessment for Federated Learning based on Cross-Round Valuation
In the federated learning (FL) process, since the data held by each participant is different, it is necessary to figure out which participant has a higher contribution to the model performance. Effective contribution assessment can help motivate data owners to participate in the FL training. Research works in this field can be divided into two directions based on whether a validation dataset is required. Validation-based methods need to use representative validation data to measure the model accuracy, which is difficult to obtain in practical FL scenarios. Existing validation-free methods assess the contribution based on the parameters and gradients of local models and the global model in a single training round, which is easily compromised by the stochasticity of model training. In this work, we propose CoAst, a practical method to assess the FL participants' contribution without access to any validation data. The core idea of CoAst involves two aspects: one is to only count the most important part of model parameters through a weights quantization, and the other is a cross-round valuation based on the similarity between the current local parameters and the global parameter updates in several subsequent communication rounds. Extensive experiments show that CoAst has comparable assessment reliability to existing validation-based methods and outperforms existing validation-free methods.
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1839-1847
Open AccessCoAst: Validation-Free Contribution Assessment for Federated Learning based on Cross-Round Valuation
In the federated learning (FL) process, since the data held by each participant is different, it is necessary to figure out which participant has a higher contribution to the model performance. Effective contribution assessment can help motivate data owners to participate in the FL training. Research works in this field can be divided into two directions based on whether a validation dataset is required. Validation-based methods need to use representative validation data to measure the model accuracy, which is difficult to obtain in practical FL scenarios. Existing validation-free methods assess the contribution based on the parameters and gradients of local models and the global model in a single training round, which is easily compromised by the stochasticity of model training. In this work, we propose CoAst, a practical method to assess the FL participants' contribution without access to any validation data. The core idea of CoAst involves two aspects: one is to only count the most important part of model parameters through a weights quantization, and the other is a cross-round valuation based on the similarity between the current local parameters and the global parameter updates in several subsequent communication rounds. Extensive experiments show that CoAst has comparable assessment reliability to existing validation-based methods and outperforms existing validation-free methods.