Zhen Xiao
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
Zhen Xiao has a total of 42 co-authors. Here are the top co-authors by number of shared publications:
- Jieyi Long
Theta Labs, Inc., San Jose, USA
7 shared publications - Rongkai Zhang
School of Computer Science, Peking University, Beijing, China
5 shared publications - Mingxuan Song
School of Computer Science, Peking University, Beijing, China
4 shared publications - Qiuyu Ding
School of Computer Science, Peking University, Beijing, China
4 shared publications - Shenglin Yin
School of Computer Science, Peking University, Beijing, China
4 shared publications - Jin Dong
School of Cyber Science and Technology, Beihang University, Beijing, China
3 shared publications
Latest Publications
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (12) , 4535-4549
Symphony: A Sharded Blockchain for Efficient Smart Contract Execution
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 75 (12) , 4535-4549
Symphony: A Sharded Blockchain for Efficient Smart Contract Execution
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 5474-5483
Alzo: Auto-Tuning with Reinforcement Learning for DAG-based Blockchains
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 5474-5483
Open AccessAlzo: Auto-Tuning with Reinforcement Learning for DAG-based Blockchains
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 40-50
DARA: Few-shot Budget Allocation in Online Advertising via In-Context Decision Making with RL-Finetuned LLMs
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 40-50
Open AccessDARA: Few-shot Budget Allocation in Online Advertising via In-Context Decision Making with RL-Finetuned LLMs
- 27 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 37 (4) , 822-835
Nexus: A Novel Transaction Processing Framework for Permissioned Blockchain
The transaction execution layer is a key determinant of throughput in permissioned blockchains. While recent Shared Memory Pools (SMP)-based approaches improve throughput by enabling all consensus nodes to participate in transaction packaging, they face two fundamental limitations. First, the performance bottleneck shifts from the consensus layer to the transaction execution layer as transaction number confirmed in a round increases. Second, these approaches are vulnerable to “transaction duplication” attacks where malicious clients can simultaneously send the same transaction to multiple consensus nodes, thereby decreasing the number of valid transactions in block proposals. To address these limitations, this paper introduces Nexus, a novel blockchain transaction processing framework with high scalability. Nexus leverages the idle computational resources of full nodes to enable transaction execution in parallel with the consensus. Moreover, Nexus allows each node to handle only a fraction of the total transactions and share execution results with others. This approach reduces overall transaction execution time, increases throughput, and decreases latency. Lastly, Nexus introduces a transaction partitioning mechanism that effectively addresses the “transaction duplication” attack and achieves load balancing between clients and consensus nodes. Our implementation of Nexus demonstrates significant improvements: throughput increases by 4x to 15x, and latency is reduced by 50% to 70%.
- 27 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 37 (4) , 822-835
Nexus: A Novel Transaction Processing Framework for Permissioned Blockchain
The transaction execution layer is a key determinant of throughput in permissioned blockchains. While recent Shared Memory Pools (SMP)-based approaches improve throughput by enabling all consensus nodes to participate in transaction packaging, they face two fundamental limitations. First, the performance bottleneck shifts from the consensus layer to the transaction execution layer as transaction number confirmed in a round increases. Second, these approaches are vulnerable to “transaction duplication” attacks where malicious clients can simultaneously send the same transaction to multiple consensus nodes, thereby decreasing the number of valid transactions in block proposals. To address these limitations, this paper introduces Nexus, a novel blockchain transaction processing framework with high scalability. Nexus leverages the idle computational resources of full nodes to enable transaction execution in parallel with the consensus. Moreover, Nexus allows each node to handle only a fraction of the total transactions and share execution results with others. This approach reduces overall transaction execution time, increases throughput, and decreases latency. Lastly, Nexus introduces a transaction partitioning mechanism that effectively addresses the “transaction duplication” attack and achieves load balancing between clients and consensus nodes. Our implementation of Nexus demonstrates significant improvements: throughput increases by 4x to 15x, and latency is reduced by 50% to 70%.
- 30 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 20 (15566013) , 8906-8921
The Halt Game: Sometimes Rewards Cannot Cover Expenses in the PoW-Based Blockchain
Proof-of-work (PoW) blockchain relies on incentive mechanisms to ensure the security and correctness of its underlying consensus protocol. Most research about it, based on a static model only considering coin-base rewards and transaction fee rewards, fails to accurately describe the complex real-world blockchain ecosystem. We propose a generic selfish mining attack applicable to arbitrary PoW blockchain systems and introduce a dynamic PoW blockchain incentive model. This model takes into account static basic rewards, dynamic whale rewards related to network protocol, and expenditures tied to players’ strategies. Unlike traditional incentive models that assume players continuously mine by default, we find players prefer to halt mining at the beginning of each mining cycle to reduce operational expenses and then resume mining at an appropriate time to enhance their rewards. We further prove players’ optimal strategy exists and it is determined by reward parameters. We implement a modified PoW blockchain system simulator and comprehensively validate these results using 256 full nodes in it of three mainstream PoW blockchains: Bitcoin, Ethereum 1.x, and Bitcoin Cash. We finally discuss the impact of different parameters on the security of PoW blockchain systems and propose practical mitigating measures for the mining halt.
- 30 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 20 (15566013) , 8906-8921
The Halt Game: Sometimes Rewards Cannot Cover Expenses in the PoW-Based Blockchain
Proof-of-work (PoW) blockchain relies on incentive mechanisms to ensure the security and correctness of its underlying consensus protocol. Most research about it, based on a static model only considering coin-base rewards and transaction fee rewards, fails to accurately describe the complex real-world blockchain ecosystem. We propose a generic selfish mining attack applicable to arbitrary PoW blockchain systems and introduce a dynamic PoW blockchain incentive model. This model takes into account static basic rewards, dynamic whale rewards related to network protocol, and expenditures tied to players’ strategies. Unlike traditional incentive models that assume players continuously mine by default, we find players prefer to halt mining at the beginning of each mining cycle to reduce operational expenses and then resume mining at an appropriate time to enhance their rewards. We further prove players’ optimal strategy exists and it is determined by reward parameters. We implement a modified PoW blockchain system simulator and comprehensively validate these results using 256 full nodes in it of three mainstream PoW blockchains: Bitcoin, Ethereum 1.x, and Bitcoin Cash. We finally discuss the impact of different parameters on the security of PoW blockchain systems and propose practical mitigating measures for the mining halt.
- Open Access
- 22 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 706-716
AERO: Enhancing Sharding Blockchain via Deep Reinforcement Learning for Account Migration
- 22 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 706-716
Open AccessAERO: Enhancing Sharding Blockchain via Deep Reinforcement Learning for Account Migration
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 541-557
Comprehensive Deadlock Prevention for GPU Collective Communication
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 541-557
Open AccessComprehensive Deadlock Prevention for GPU Collective Communication
- Open Access
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 9435-9444
Embracing Adaptation: An Effective Dynamic Defense Strategy Against Adversarial Examples
Existing adversarial example defense methods are static, meaning they remain unchanged once training is completed, regardless of how attack methods change. Consequently, static defense methods are highly vulnerable to adaptive attacks. We argue that to counter more formidable attacks, models should continually adapt to various attack methods. We propose a novel dynamic defense approach. Initially, we use Gaussian Mixture Models (GMM) to obtain structural information of the data, which is combined with model prediction information to generate pseudo-labels for optimizing inputs. Subsequently, we employ information maximization and enhanced mean predictions as optimization objectives, utilizing a hierarchical optimization approach to refine the model. Meanwhile, we propose a sample-efficient optimization strategy that reduces the total number of samples in the test data stream for reverse updating and improves the efficiency. Notably, our method can be directly applied to pre-trained models without the need for accessing training data or retraining the model. Therefore, our approach is training-data-agnostic and model-agnostic, easily applicable to existing adversarially trained models, significantly enhancing the resilience of various models against white-box, black-box, and adaptive attacks across diverse datasets. We have conducted extensive experiments to validate the state-of-the-art of our proposed method. The pseudo-code can be found in the appendix.
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 9435-9444
Open AccessEmbracing Adaptation: An Effective Dynamic Defense Strategy Against Adversarial Examples
Existing adversarial example defense methods are static, meaning they remain unchanged once training is completed, regardless of how attack methods change. Consequently, static defense methods are highly vulnerable to adaptive attacks. We argue that to counter more formidable attacks, models should continually adapt to various attack methods. We propose a novel dynamic defense approach. Initially, we use Gaussian Mixture Models (GMM) to obtain structural information of the data, which is combined with model prediction information to generate pseudo-labels for optimizing inputs. Subsequently, we employ information maximization and enhanced mean predictions as optimization objectives, utilizing a hierarchical optimization approach to refine the model. Meanwhile, we propose a sample-efficient optimization strategy that reduces the total number of samples in the test data stream for reverse updating and improves the efficiency. Notably, our method can be directly applied to pre-trained models without the need for accessing training data or retraining the model. Therefore, our approach is training-data-agnostic and model-agnostic, easily applicable to existing adversarially trained models, significantly enhancing the resilience of various models against white-box, black-box, and adaptive attacks across diverse datasets. We have conducted extensive experiments to validate the state-of-the-art of our proposed method. The pseudo-code can be found in the appendix.
- Open Access
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 6061-6072
Understanding the Weakness of Large Language Model Agents within a Complex Android Environment
Large language models (LLMs) have empowered intelligent agents to execute intricate tasks within domain-specific software such as browsers and games. However, when applied to general-purpose software systems like operating systems, LLM agents face three primary challenges. Firstly, the action space is vast and dynamic, posing difficulties for LLM agents to maintain an up-to-date understanding and deliver accurate responses. Secondly, real-world tasks often require inter-application cooperation, demanding farsighted planning from LLM agents. Thirdly, agents need to identify optimal solutions aligning with user constraints, such as security concerns and preferences. These challenges motivate AndroidArena, an environment and benchmark designed to evaluate LLM agents on a modern operating system. To address high-cost of manpower, we design a scalable and semi-automated method to construct the benchmark. In the task evaluation, AndroidArena incorporates accurate and adaptive metrics to address the issue of non-unique solutions. Our findings reveal that even state-of-the-art LLM agents struggle in cross-APP scenarios and adhering to specific constraints. Additionally, we identify a lack of four key capabilities, i.e. understanding, reasoning, exploration, and reflection, as primary reasons for the failure of LLM agents. Furthermore, we provide empirical analysis on the failure of reflection, and improve the success rate by 27% with our proposed exploration strategy. This work is the first to present valuable insights in understanding fine-grained weakness of LLM agents, and offers a path forward for future research in this area. Environment, benchmark, prompt, and evaluation code for AndroidArena are released at https://github.com/AndroidArenaAgent/AndroidArena.
- 24 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 6061-6072
Open AccessUnderstanding the Weakness of Large Language Model Agents within a Complex Android Environment
Large language models (LLMs) have empowered intelligent agents to execute intricate tasks within domain-specific software such as browsers and games. However, when applied to general-purpose software systems like operating systems, LLM agents face three primary challenges. Firstly, the action space is vast and dynamic, posing difficulties for LLM agents to maintain an up-to-date understanding and deliver accurate responses. Secondly, real-world tasks often require inter-application cooperation, demanding farsighted planning from LLM agents. Thirdly, agents need to identify optimal solutions aligning with user constraints, such as security concerns and preferences. These challenges motivate AndroidArena, an environment and benchmark designed to evaluate LLM agents on a modern operating system. To address high-cost of manpower, we design a scalable and semi-automated method to construct the benchmark. In the task evaluation, AndroidArena incorporates accurate and adaptive metrics to address the issue of non-unique solutions. Our findings reveal that even state-of-the-art LLM agents struggle in cross-APP scenarios and adhering to specific constraints. Additionally, we identify a lack of four key capabilities, i.e. understanding, reasoning, exploration, and reflection, as primary reasons for the failure of LLM agents. Furthermore, we provide empirical analysis on the failure of reflection, and improve the success rate by 27% with our proposed exploration strategy. This work is the first to present valuable insights in understanding fine-grained weakness of LLM agents, and offers a path forward for future research in this area. Environment, benchmark, prompt, and evaluation code for AndroidArena are released at https://github.com/AndroidArenaAgent/AndroidArena.
- Open Access
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 2836-2846
SPRING: Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement
- 13 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 2836-2846
Open AccessSPRING: Improving the Throughput of Sharding Blockchain via Deep Reinforcement Learning Based State Placement