Adaptive sharding via transformer-enhanced reinforcement learning for blockchain-based edge computing

The growing scale of data-intensive applications and resource-constrained devices in mobile edge computing (MEC) demands both secure data processing and high system throughput. Traditional blockchains enhance trust and immutability but suffer from low transaction rates and high confirmation latency, making them ill-suited for real-time edge scenarios. This work addresses the challenge of jointly satisfying security and performance requirements in blockchain-enabled MEC. We target an MEC environment where edge nodes cooperate in processing and validating tasks under dynamic network conditions. We develop an adaptive sharding mechanism, which partitions the network into preliminary shards of varying configurations, and design a deep reinforcement learning algorithm to maximize system utility under a provable security threshold. Expanding from the conference version of this paper[1], we propose a novel Transformer-enhanced model and implement a working prototype on the Cosmos SDK and demonstrate cross-chain consensus and task offloading under real MEC settings. Experimental evaluations show that our adaptive sharding scheme significantly improves transaction throughput compared to baselines while maintaining security guarantees.

Authors

Institutions

Publication Details

Journal
Complex & Intelligent Systems
Published
2026-11-16
DOI
https://doi.org/10.1007/s40747-026-02437-7
Primary Topic
Blockchain Technology Applications and Security
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Adaptive sharding via transformer-enhanced reinforcement learning for blockchain-based edge computing

Shijing Yuan, Jie Li, Suchuan Xing, Hongyang Chen et al.
1 links
Complex & Intelligent Systems
Blockchain Technology Applications and Security
article

Adaptive sharding via transformer-enhanced reinforcement learning for blockchain-based edge computing

Shijing Yuan, Jie Li, Suchuan Xing, Hongyang Chen, Zhi Liu, Xiang Chen, Athanasios V. Vasilakos, Wei Zhao
article en

Abstract

The growing scale of data-intensive applications and resource-constrained devices in mobile edge computing (MEC) demands both secure data processing and high system throughput. Traditional blockchains enhance trust and immutability but suffer from low transaction rates and high confirmation latency, making them ill-suited for real-time edge scenarios. This work addresses the challenge of jointly satisfying security and performance requirements in blockchain-enabled MEC. We target an MEC environment where edge nodes cooperate in processing and validating tasks under dynamic network conditions. We develop an adaptive sharding mechanism, which partitions the network into preliminary shards of varying configurations, and design a deep reinforcement learning algorithm to maximize system utility under a provable security threshold. Expanding from the conference version of this paper[1], we propose a novel Transformer-enhanced model and implement a working prototype on the Cosmos SDK and demonstrate cross-chain consensus and task offloading under real MEC settings. Experimental evaluations show that our adaptive sharding scheme significantly improves transaction throughput compared to baselines while maintaining security guarantees.

Complex & Intelligent Systems
Duke University (US), Shanghai Jiao Tong University (CN), Chinese Academy of Sciences (CN), University of Agder (NO), University of Electro-Communications (JP), Zhejiang Lab (CN), China Telecom (China) (CN), Shenzhen Institutes of Advanced Technology (CN)
Industry, innovation and infrastructure
Blockchain Technology Applications and Security
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon