He Sun
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
He Sun has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Mingjun Xiao
School of Computer Science and Technology, University of Science and Technology, Hefei, China
8 shared publications - Jinrui Zhou
University of Science and Technology of China Hefei China
4 shared publications - Yin Xu
University of Science and Technology of China
4 shared publications - Jie Wu
Cloud Computing Research Institute, China Telecom
3 shared publications - Li Li
University of Macau, Macau, Macao
2 shared publications - Chen Wu
School of Data Science, University of Science and Technology of China, Hefei, China
1 shared publication
Latest Publications
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4463-4474
HeimdaLLM: Efficient Cloud-assisted Federated Fine-tuning with Zeroth-Order Rectification for LLMs
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4463-4474
Open AccessHeimdaLLM: Efficient Cloud-assisted Federated Fine-tuning with Zeroth-Order Rectification for LLMs
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (7) , 10211-10224
Edge-Assisted Multi-Vehicle Cooperative Perception: An Approach Based on Relative Pose Estimations
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (7) , 10211-10224
Edge-Assisted Multi-Vehicle Cooperative Perception: An Approach Based on Relative Pose Estimations
- Open Access
CooperLLM: Cloud-Edge-End Cooperative Federated Fine-tuning for LLMs via ZOO-based Gradient Correction
Large Language Models (LLMs) perform well on many NLP tasks, but fine-tuning them on resource-constrained mobile devices is challenging due to high memory and computation costs, despite growing demands for privacy-preserving personalization. Federated Learning (FL) enables local-data training, yet existing methods either rely on memory-intensive backpropagation or use zeroth-order optimization (ZOO), which avoids backward passes but suffers from slow convergence and degraded accuracy. We propose CooperLLM, a cloud-assisted edge-end cooperative federated fine-tuning framework that combines ZOO on mobile devices with cloud-guided gradient rectification. Mobile clients perform lightweight ZOO updates on private data, while the cloud fine-tunes on auxiliary public data using backpropagation and injects guided perturbations to rectify local updates, improving convergence and accuracy without violating privacy. To address system bottlenecks, CooperLLM introduces pipeline scheduling and adaptive compression to overlap computation and communication and reduce memory usage. Experiments on multiple Transformer models and datasets show that CooperLLM reduces on-device memory by up to $86.4\%$, accelerates convergence by $8.8 \times$, and improves accuracy by up to 10 percentage points over state-of-the-art ZOO-based baselines.Open AccessCooperLLM: Cloud-Edge-End Cooperative Federated Fine-tuning for LLMs via ZOO-based Gradient Correction
Large Language Models (LLMs) perform well on many NLP tasks, but fine-tuning them on resource-constrained mobile devices is challenging due to high memory and computation costs, despite growing demands for privacy-preserving personalization. Federated Learning (FL) enables local-data training, yet existing methods either rely on memory-intensive backpropagation or use zeroth-order optimization (ZOO), which avoids backward passes but suffers from slow convergence and degraded accuracy. We propose CooperLLM, a cloud-assisted edge-end cooperative federated fine-tuning framework that combines ZOO on mobile devices with cloud-guided gradient rectification. Mobile clients perform lightweight ZOO updates on private data, while the cloud fine-tunes on auxiliary public data using backpropagation and injects guided perturbations to rectify local updates, improving convergence and accuracy without violating privacy. To address system bottlenecks, CooperLLM introduces pipeline scheduling and adaptive compression to overlap computation and communication and reduce memory usage. Experiments on multiple Transformer models and datasets show that CooperLLM reduces on-device memory by up to $86.4\%$, accelerates convergence by $8.8 \times$, and improves accuracy by up to 10 percentage points over state-of-the-art ZOO-based baselines. - Open Access
SFedKD: Sequential Federated Learning with Discrepancy-Aware Multi-Teacher Knowledge Distillation
Federated Learning (FL) is a distributed machine learning paradigm which coordinates multiple clients to collaboratively train a global model via a central server. Sequential Federated Learning (SFL) is a newly-emerging FL training framework where the global model is trained in a sequential manner across clients. Since SFL can provide strong convergence guarantees under data heterogeneity, it has attracted significant research attention in recent years. However, experiments show that SFL suffers from severe catastrophic forgetting in heterogeneous environments, meaning that the model tends to forget knowledge learned from previous clients. To address this issue, we propose an SFL framework with discrepancy-aware multi-teacher knowledge distillation, called SFedKD, which selects multiple models from the previous round to guide the current round of training. In SFedKD, we extend the single-teacher Decoupled Knowledge Distillation approach to our multi-teacher setting and assign distinct weights to teachers' target-class and non-target-class knowledge based on the class distributional discrepancy between teacher and student data. Through this fine-grained weighting strategy, SFedKD can enhance model training efficacy while mitigating catastrophic forgetting. Additionally, to prevent knowledge dilution, we eliminate redundant teachers for the knowledge distillation and formalize it as a variant of the maximum coverage problem. Based on the greedy strategy, we design a complementary-based teacher selection mechanism to ensure that the selected teachers achieve comprehensive knowledge space coverage while reducing communication and computational costs. Extensive experiments show that SFedKD effectively overcomes catastrophic forgetting in SFL and outperforms state-of-the-art FL methods.
Open AccessSFedKD: Sequential Federated Learning with Discrepancy-Aware Multi-Teacher Knowledge Distillation
Federated Learning (FL) is a distributed machine learning paradigm which coordinates multiple clients to collaboratively train a global model via a central server. Sequential Federated Learning (SFL) is a newly-emerging FL training framework where the global model is trained in a sequential manner across clients. Since SFL can provide strong convergence guarantees under data heterogeneity, it has attracted significant research attention in recent years. However, experiments show that SFL suffers from severe catastrophic forgetting in heterogeneous environments, meaning that the model tends to forget knowledge learned from previous clients. To address this issue, we propose an SFL framework with discrepancy-aware multi-teacher knowledge distillation, called SFedKD, which selects multiple models from the previous round to guide the current round of training. In SFedKD, we extend the single-teacher Decoupled Knowledge Distillation approach to our multi-teacher setting and assign distinct weights to teachers' target-class and non-target-class knowledge based on the class distributional discrepancy between teacher and student data. Through this fine-grained weighting strategy, SFedKD can enhance model training efficacy while mitigating catastrophic forgetting. Additionally, to prevent knowledge dilution, we eliminate redundant teachers for the knowledge distillation and formalize it as a variant of the maximum coverage problem. Based on the greedy strategy, we design a complementary-based teacher selection mechanism to ensure that the selected teachers achieve comprehensive knowledge space coverage while reducing communication and computational costs. Extensive experiments show that SFedKD effectively overcomes catastrophic forgetting in SFL and outperforms state-of-the-art FL methods.
- 29 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 35 (11) , 2039-2053
Privacy Preserving Task Push in Spatial Crowdsourcing With Unknown Popularity
In this paper, we investigate the privacy-preserving task push problem with unknown popularity in Spatial Crowdsourcing (SC), where the platform needs to select some tasks with unknown popularity and push them to workers. Meanwhile, the preferences of workers and the popularity values of tasks might involve some sensitive information, which should be protected from disclosure. To address these concerns, we propose a Privacy Preserving Auction-based Bandit scheme, termed PPAB. Specifically, on the basis of the Combinatorial Multi-armed Bandit (CMAB) game, we first construct a Differentially Private Auction-based CMAB (DPA-CMAB) model. Under the DPA-CMAB model, we design a privacy-preserving arm-pulling policy based on Diffie-Hellman (DH), Differential Privacy (DP), and upper confidence bound, which includes the DH-based encryption mechanism and the hybrid DP-based protection mechanism. The policy not only can learn the popularity of tasks and make online task push decisions, but also can protect the popularity as well as workers’ preferences from being revealed. Meanwhile, we design an auction-based incentive mechanism to determine the payment for each selected task. Furthermore, we conduct an in-depth analysis of the security and online performance of PPAB, and prove that PPAB satisfies some desired properties (i.e., truthfulness, individual rationality, and computational efficiency). Finally, the significant performance of PPAB is confirmed through extensive simulations on the real-world dataset.
- 29 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 35 (11) , 2039-2053
Privacy Preserving Task Push in Spatial Crowdsourcing With Unknown Popularity
In this paper, we investigate the privacy-preserving task push problem with unknown popularity in Spatial Crowdsourcing (SC), where the platform needs to select some tasks with unknown popularity and push them to workers. Meanwhile, the preferences of workers and the popularity values of tasks might involve some sensitive information, which should be protected from disclosure. To address these concerns, we propose a Privacy Preserving Auction-based Bandit scheme, termed PPAB. Specifically, on the basis of the Combinatorial Multi-armed Bandit (CMAB) game, we first construct a Differentially Private Auction-based CMAB (DPA-CMAB) model. Under the DPA-CMAB model, we design a privacy-preserving arm-pulling policy based on Diffie-Hellman (DH), Differential Privacy (DP), and upper confidence bound, which includes the DH-based encryption mechanism and the hybrid DP-based protection mechanism. The policy not only can learn the popularity of tasks and make online task push decisions, but also can protect the popularity as well as workers’ preferences from being revealed. Meanwhile, we design an auction-based incentive mechanism to determine the payment for each selected task. Furthermore, we conduct an in-depth analysis of the security and online performance of PPAB, and prove that PPAB satisfies some desired properties (i.e., truthfulness, individual rationality, and computational efficiency). Finally, the significant performance of PPAB is confirmed through extensive simulations on the real-world dataset.
- 13 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 23 (12) , 11719-11734
Crowdsensing Data Trading for Unknown Market: Privacy, Stability, and Conflicts
In recent years, Crowdsensing Data Trading (CDT) has emerged as a new data trading paradigm, where buyers crowdsource data collection tasks to a group of mobile users with sensing devices (a.k.a., sellers) who sell the collected data to them, through a platform as the broker for a long-term data trading. One of the most critical issues in CDT is ensuring the stability of the matching between buyers and sellers in the data trading market. In this paper, we focus on privacy protection and the stability problems in the CDT market with unknown preference sequences of buyers. The goal is to protect sellers' data qualities and ensure the CDT market's stability while maximizing the cumulative data quality for each task. We model this problem as a differentially private multi-player multi-armed competing bandit problem and propose a novel metric of the approximate stability, called $\delta$ -stability. We propose a privacy-preserving stable CDT mechanism called DPS-CB to solve this problem in the centralized setting, which is based on stable matching theory, and competing bandit strategy. Moreover, we extend it into decentralized setting in order to avoid the competitive matching conflicts caused in this setting and propose a Conflicts-avoiding DPS-CB mechanism, called CDPS-CB, by using Bernoulli probability and selecting feasible sets of sellers. In addition, we prove the security and stability of the CDT market under privacy concerns and analyze the regret performance of DPS-CB and CDPS-CB mechanisms, respectively. Finally, the significant performance of these two mechanisms is demonstrated through extensive simulations on a real-world dataset.
- 13 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 23 (12) , 11719-11734
Crowdsensing Data Trading for Unknown Market: Privacy, Stability, and Conflicts
In recent years, Crowdsensing Data Trading (CDT) has emerged as a new data trading paradigm, where buyers crowdsource data collection tasks to a group of mobile users with sensing devices (a.k.a., sellers) who sell the collected data to them, through a platform as the broker for a long-term data trading. One of the most critical issues in CDT is ensuring the stability of the matching between buyers and sellers in the data trading market. In this paper, we focus on privacy protection and the stability problems in the CDT market with unknown preference sequences of buyers. The goal is to protect sellers' data qualities and ensure the CDT market's stability while maximizing the cumulative data quality for each task. We model this problem as a differentially private multi-player multi-armed competing bandit problem and propose a novel metric of the approximate stability, called $\delta$ -stability. We propose a privacy-preserving stable CDT mechanism called DPS-CB to solve this problem in the centralized setting, which is based on stable matching theory, and competing bandit strategy. Moreover, we extend it into decentralized setting in order to avoid the competitive matching conflicts caused in this setting and propose a Conflicts-avoiding DPS-CB mechanism, called CDPS-CB, by using Bernoulli probability and selecting feasible sets of sellers. In addition, we prove the security and stability of the CDT market under privacy concerns and analyze the regret performance of DPS-CB and CDPS-CB mechanisms, respectively. Finally, the significant performance of these two mechanisms is demonstrated through extensive simulations on a real-world dataset.
- 1 May 2024
- Published by Springer Nature in Journal of Computer Science and Technology
- Vol. 39 (3) , 637-653
Age-of-Information-Aware Federated Learning
- 1 May 2024
- Published by Springer Nature in Journal of Computer Science and Technology
- Vol. 39 (3) , 637-653
Age-of-Information-Aware Federated Learning