.png){kind=icon}
.png){kind=icon}
Return
Dynamic transaction scheduling for data elements: A multi-level queueing model with LSTM-based load prediction
.png){kind=icon}
OA
AI
DOI：10.1016/j.hcc.2026.100400.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The emergence of data as a critical factor of production necessitates efficient and trusted markets for its exchange. Blockchain technology, serving as a key infrastructure for such data trading platforms, faces severe transaction congestion. This issue is exacerbated by the conventional ”fee-first” mechanism, which fails to account for the heterogeneous time-sensitivity of data transactions, leading to inefficient resource allocation and unfair delays that undermine the value of time-critical data. To address this, we propose a dynamic scheduling framework tailored for blockchain-based data trading. Our solution integrates a non-preemptive multi-priority queueing system based on the M/M/1 model with a Long Short-Term Memory (LSTM) network for load prediction. The framework categorizes data transaction requests into distinct priority queues based on their urgency, employs the LSTM to predict transaction arrival rates for real-time parameter adjustment, and incorporates a theoretical analysis for performance estimation. Extensive experiments using synthetic datasets (simulating various market conditions via Poisson, normal, and exponential distributions) and real Ethereum transaction data demonstrate that our multi-queue mechanism consistently outperforms traditional single-queue systems. It effectively reduces waiting times for high-priority data transactions, lowers associated costs, and maintains robust performance under high-frequency fluctuations and extreme congestion. This work provides a theoretically grounded and practical solution for enhancing transaction efficiency, fairness, and user experience in blockchain-based data element markets.
Keywords:
Blockchain
Multi-level queueing
M/M/1 model
LSTM
Transaction scheduling
Dynamic load prediction
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
H
IF:
3
Papers:
239
Citations:
407
View more organizations