Areas of Interest

Blockchain Interoperability
Cross-chain technology facilitates the interaction between two independent blockchains. Depending on the type of cross-chain activities, cross-chain interactions can be divided into cross-chain transfers and cross-chain exchanges. Cross-chain transfers refer to the movement of assets between two blockchains, which involve the process of burning assets on one blockchain and minting them on the other. Cross-chain exchanges, conceptually, involve a two-way process of cross-chain transfers. They refer to the swapping of assets between two distinct blockchains. Our research is dedicated to solving various problems posed by these different activities, including but not limited to those concerning efficiency, security, and privacy.
[USENIX Security'24] zkCross: A Novel Architecture for Cross-Chain Privacy-Preserving Auditing
[ACM Computing Surveys'24] Exploring Blockchain Technology through a Modular Lens: A Survey
[IEEE JSAC'23] An Adaptive and Modular Blockchain Enabled Architecture for a Decentralized Metaverse
[IEEE TC'23] Cross-Channel: Scalable Off-Chain Channels Supporting Fair and Atomic Cross-Chain Operations

Decentralized Storage and Computing
Decentralized storage networks aggregate free storage spaces offered by independent storage providers and self-coordinate to provide data storage and retrieval services. Compared to traditional storage networks, a decentralized storage network is operated on a blockchain system, which works as an incentive layer. Blockchain rewards storage providers who provide reliable storage to clients, and thus enables an open manageable storage market. Decentralized computing utilizes decentralized networks of independent nodes to perform computation collaboratively without relying on centralized infrastructure. Coordinated by blockchain, these systems ensure transparent task allocation and fair rewards, creating an open and reliable market for computing power with enhanced scalability and resilience.
[IEEE INFOCOM'25] Paper Title
[IEEE INFOCOM'24] FileDES: A Secure, Scalable and Succinct Blockchain-based Decentralized Encrypted Storage Network
[IEEE TC'24] BFT-DSN: A Byzantine Fault Tolerant Decentralized Storage Network
[IEEE TC'23] FileDAG: A Multi-Version Decentralized Storage Network Built on DAG-based Blockchain
[IEEE ICDCS'22] Curb: Trusted and Scalable Software-Defined Network Control Plane for Edge Computing
[IEEE TMC'23] TBAC: A Tokoin-based Accountable Access Control Scheme for the Internet of Things
[IEEE TC'22] Extending On-chain Trust to Off-chain--Trustworthy Blockchain Data Collection using Trusted Execution Environment (TEE)

Byzantine Fault-Tolerant Consensus
Byzantine fault-tolerant (BFT) consensus is a method for achieving consensus or agreement within a distributed network, even if some of the nodes in the network are unreliable or malicious. The method is named after the Byzantine Generals’ Problem, a hypothetical scenario in which a group of generals must coordinate their attack on a city. In the problem, some of the generals may be traitors who will attempt to sabotage the attack. In the context of distributed networks, the problem refers to the challenge of achieving consensus among nodes that may be compromised or malfunctioning. BFT consensus algorithms aim to ensure that the network as a whole arrives at a consistent state, even if individual nodes are acting in bad faith or providing unreliable information. This is achieved through sophisticated cryptographic techniques and mathematical algorithms. By using BFT consensus, distributed networks can achieve a high level of trust and resilience, making them well-suited for applications such as blockchain technology.
[IEEE INFOCOM'25] Partially Synchronous BFT Consensus Made Practical in Wireless Networks
[IEEE ICDCS'25] Asynchronous BFT Consensus Made Wireless
[IEEE TMC'22] BLOWN: A Blockchain Protocol for Single-Hop Wireless Networks under Adversarial SINR
[IEEE TC'22] CloudChain: A Cloud Blockchain Using Shared Memory Consensus and RDMA
[IEEE TWC'21] wChain: A Fast Fault-Tolerant Blockchain Protocol for Multihop Wireless Networks
[arXiv] Sleepy Consensus in the Known Participation Model

Applied Cryptography
Applied Cryptography is the discipline of using cryptographic techniques to build secure systems and solve real-world security problems. It focuses on the practical implementation and application of cryptographic algorithms and protocols. The core challenge in this field is to protect information and communications in the presence of adversaries who may attempt to access, modify, or disrupt data. Applied Cryptography addresses threats such as eavesdropping, data tampering, unauthorized access, and denial of service. This is achieved by employing a variety of cryptographic tools, including encryption algorithms, digital signatures, hash functions, and key management techniques, often combined into complex security protocols. By using Applied Cryptography, systems can achieve confidentiality, integrity, authenticity, and non-repudiation, making it essential for applications such as secure communication, data storage, electronic commerce, and digital identity.
[IEEE TIFS'25] Trinity: A Scalable and Forward-Secure DSSE for Spatio-Temporal Range Query
[IEEE INFOCOM'23] Latency-First Smart Contract: Overclock the Blockchain for a While
[IEEE TC'22] Split: A Hash-based Memory Optimization Method for Zero-Knowledge Succinct Non-Interactive Argument of Knowledge (zk-SNARK)
[arXiv] VDORAM: Towards a Random Access Machine with Both Public Verifiability and Distributed Obliviousness
[arXiv] AD-MPC: Fully Asynchronous Dynamic MPC with Guaranteed Output Delivery

AI (LLM) Security & Privacy
AI (LLM) security and privacy focus on safeguarding large language models and their usage from risks such as data leakage, model inversion, prompt injection, and unauthorized access. As LLMs often interact with sensitive inputs and generate influential outputs, ensuring robust access control, secure data handling, and adversarial resistance is critical. Privacy-preserving techniques like differential privacy, secure multi-party computation, and federated learning are increasingly applied to protect user data and model integrity in both training and deployment phases.
[IEEE ICMC'24] On Protecting the Data Privacy of Large Language Models (LLMs): A Survey
[IEEE TC'24] FedRFQ: Prototype-based Federated Learning with Reduced Redundancy, Minimal Failure, and Enhanced Quality
[IEEE TC'22] SPDL: A Blockchain-enabled Secure and Privacy-preserving Decentralized Learning System
[arXiv] I'm Spartacus, No, I'm Spartacus: Measuring and Understanding LLM Identity Confusion