Sheng Cao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Sheng Cao has a total of 65 co-authors. Here are the top co-authors by number of shared publications:
- Xiaosong Zhang
University of Electronic Science and Technology of China
11 shared publications - Man Ho Au
The Hong Kong Polytechnic University
3 shared publications - Xiaofen Wang
University of Electronic Science and Technology of China
3 shared publications Xiaosong Zhang
University of Electronic Science and Technology of China
3 shared publications- Butian Huang
Hangzhou Dianzi University
2 shared publications - Hao Huang
School of Computer Science and Engineering (School of Cyber Security), University of Electronic Science and Technology of China, Chengdu, China
2 shared publications
Latest Publications
- 4 November 2026
- Published by Springer Nature
- p. 547-574
Scaling Blockchains Under a Dishonest Majority
- Open Access
- 17 October 2026
- Published by Elsevier in Digital Communications and Networks
- Vol. 12 (10) , 1698-1708
End-to-end robust H.265 video watermarking over lossy channels: a differentiable encoding and spatio-temporal feature synergy framework
- 17 October 2026
- Published by Elsevier in Digital Communications and Networks
- Vol. 12 (10) , 1698-1708
Open AccessEnd-to-end robust H.265 video watermarking over lossy channels: a differentiable encoding and spatio-temporal feature synergy framework
- 8 October 2026
- Published by Elsevier in Information Sciences
- Vol. 762
A dynamic graph query-based data pricing method via multi-armed bandits
- 8 October 2026
- Published by Elsevier in Information Sciences
- Vol. 762
A dynamic graph query-based data pricing method via multi-armed bandits
- 1 August 2026
- Published by Springer Nature
- p. 371-387
Case-Based Calibration of Adaptive Reasoning and Execution for LLM Tool Use
- 1 August 2026
- Published by Springer Nature
- p. 371-387
Case-Based Calibration of Adaptive Reasoning and Execution for LLM Tool Use
- 12 July 2026
- Published by Springer Nature
- p. 529-543
Dynamic Pricing Method via Multi-armed Bandits in Graph Query-Based Data Market
- 12 July 2026
- Published by Springer Nature
- p. 529-543
Dynamic Pricing Method via Multi-armed Bandits in Graph Query-Based Data Market
- 24 April 2026
- Published by Elsevier in Information Fusion
- Vol. 135
EMNet: Efficient Multimodal Network for Action Recognition
- 24 April 2026
- Published by Elsevier in Information Fusion
- Vol. 135
EMNet: Efficient Multimodal Network for Action Recognition
- 2 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21 (15566013) , 3615-3629
CShard: Blockchain Sharding via Repairable Fountain Codes and the Paradigm for Sharding Design
Sharding is an important solution to improve the scalability of blockchain. The basic idea of blockchain sharding is to separate transactions among multiple disjoint shards processing in parallel to maximize system performance. The current sharding protocols mainly rely on node rotation (namely, node allocation and migration) randomly among shards periodically to ensure security, which is often considered the most challenging when developing a sharding system. However, (1) if a shard or multiple shards are corrupted, the sharding system will not be available anymore; (2) to avoid shard corruption, the demand of large size of each shard limits the throughput performance of sharding protocols. To solve (1), we introduce a blockchain sharding protocol with scale-out transaction processing capacity called CShard. The main idea of CShard is to use repairable fountain codes (RFCs), an information coding method with the locality feature, to innovate sharding design. By adjusting encoding parameters, topological associations among shards are constructed, which are then utilized to define the verification logic for transactions. The blocks of corrupted shard(s) can be recovered through decoding by its corresponding shard group(s), and the sharding system is still secure and available. Our approach utilizes encoding techniques to build a general architecture of a sharding system, establishing a paradigm of “encoding as verification” and showcases new horizons in the field of blockchain sharding. To solve (2), we propose the ghost reporter mechanism that gives all nodes chances to verify a transaction by submitting reports in the sharding network. The mechanism brings two direct benefits. Firstly, it provides the way to detect corrupted shards and recover the blocks by RFCs; the second is to make the number of nodes in a single shard smaller, which solves the limitation of existing sharding schemes that usually require a larger number of nodes in a single shard to ensure security. In principle, this mechanism can also be applicable to the known and even unknown sharding protocols for its generality.
- 2 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21 (15566013) , 3615-3629
CShard: Blockchain Sharding via Repairable Fountain Codes and the Paradigm for Sharding Design
Sharding is an important solution to improve the scalability of blockchain. The basic idea of blockchain sharding is to separate transactions among multiple disjoint shards processing in parallel to maximize system performance. The current sharding protocols mainly rely on node rotation (namely, node allocation and migration) randomly among shards periodically to ensure security, which is often considered the most challenging when developing a sharding system. However, (1) if a shard or multiple shards are corrupted, the sharding system will not be available anymore; (2) to avoid shard corruption, the demand of large size of each shard limits the throughput performance of sharding protocols. To solve (1), we introduce a blockchain sharding protocol with scale-out transaction processing capacity called CShard. The main idea of CShard is to use repairable fountain codes (RFCs), an information coding method with the locality feature, to innovate sharding design. By adjusting encoding parameters, topological associations among shards are constructed, which are then utilized to define the verification logic for transactions. The blocks of corrupted shard(s) can be recovered through decoding by its corresponding shard group(s), and the sharding system is still secure and available. Our approach utilizes encoding techniques to build a general architecture of a sharding system, establishing a paradigm of “encoding as verification” and showcases new horizons in the field of blockchain sharding. To solve (2), we propose the ghost reporter mechanism that gives all nodes chances to verify a transaction by submitting reports in the sharding network. The mechanism brings two direct benefits. Firstly, it provides the way to detect corrupted shards and recover the blocks by RFCs; the second is to make the number of nodes in a single shard smaller, which solves the limitation of existing sharding schemes that usually require a larger number of nodes in a single shard to ensure security. In principle, this mechanism can also be applicable to the known and even unknown sharding protocols for its generality.
- 2 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 13, 4107-4123
GBShard: Adaptive Sharding for Blockchain Network via Granular-Ball Partitioning
Blockchain scalability remains constrained by cross-shard transaction overhead and workload imbalance in conventional sharding architectures. To address these limitations, we propose GBShard, an adaptive sharding method grounded in granular-ball computing theory. The method partitions blockchain transactions to shards through iterative splitting and merging of coarse-grained granular-ball structures, enabling adaptive shard formation with minimized cross-shard transactions. A multi-granularity dynamic sharding algorithm further optimizes topology by redistributing granular-ball groups in dynamic scenarios, achieving incremental sharding adjustment while reducing the overhead of global-repartitioning. Experimental validation using real-world Ethereum transaction traces demonstrates GBShard’s superiority over Monoxide-based sharding schemes: it achieves 1.3–1.8× higher throughput, 34–42% lower transaction latency, and reduces cross-shard transactions by 23–50%. These results suggest granular-ball partitioning as a viable strategy for adaptive blockchain scaling.
- 2 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 13, 4107-4123
GBShard: Adaptive Sharding for Blockchain Network via Granular-Ball Partitioning
Blockchain scalability remains constrained by cross-shard transaction overhead and workload imbalance in conventional sharding architectures. To address these limitations, we propose GBShard, an adaptive sharding method grounded in granular-ball computing theory. The method partitions blockchain transactions to shards through iterative splitting and merging of coarse-grained granular-ball structures, enabling adaptive shard formation with minimized cross-shard transactions. A multi-granularity dynamic sharding algorithm further optimizes topology by redistributing granular-ball groups in dynamic scenarios, achieving incremental sharding adjustment while reducing the overhead of global-repartitioning. Experimental validation using real-world Ethereum transaction traces demonstrates GBShard’s superiority over Monoxide-based sharding schemes: it achieves 1.3–1.8× higher throughput, 34–42% lower transaction latency, and reduces cross-shard transactions by 23–50%. These results suggest granular-ball partitioning as a viable strategy for adaptive blockchain scaling.