An Intelligent Trust Prediction and Adaptive Data Placement Framework for Secure Cloud–Edge Data Sharing
DOI:
https://doi.org/10.64764/xxfaxr61Keywords:
Cloud–Edge Computing; Secure Data Distribution; Trust Prediction; Machine Learning; Adaptive Data Placement; Trust Management; Erasure Coding; Fault Tolerance; Data Security; Edge Computing.Abstract
Cloud-edge computing has undergone significant growth, adding to the growing demand for security, reliability, and adaptive solutions to distribute sensitive data across diverse and often untrustworthy edge nodes. Traditional approaches to data placement mainly rely on static data placement policies and/or availability of resources, potentially allowing for exposure of data to unreliable and/or compromised nodes. In order to solve this problem, in this paper, an innovative secure framework of Intelligent Trust Prediction and Adaptive Data Placement (ITADP-Edge) for intelligent data distribution in the cloud–edge environment is presented. This corresponds to behavioral monitoring, trust prediction via machine learning, adaptive node selection in order to ensure low-latency networking, encryption, erasure coding, integrity verification, and fault recovery being compiled in one security framework. The trust-prediction component was tested by the CICIDS2017 dataset and was modeled with Decision Tree, Random Forest, XGBoost, and LightGBM models for separating benign and potentially malicious behavior. The proposed LightGBM-based ITADP-Edge achieved 99.84% accuracy, 99.73% precision, 99.83% recall, 99.78% F1-score, and 0.9999 ROC-AUC. Moreover, the framework adds trust information to the nodes' ranking based on their predicted trust values, allowing the nodes to place data more securely and adaptively instead of statically distributing data. Encryption helps to provide data confidentiality, and erasure coding increases resistance to the failure of nodes and data integrity. The proposed framework gives strong support to secure, trustworthy, and fault-tolerant cloud-edge distribution of both data and information with low latency.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Author(s)

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.






