Blockchain-Enabled Secure Data Sharing in Cloudedge Learning Networks

Abstract

Cloud edge convergence is enabling real time personalized learning but come with security and trust issues. Centralized models cause single points of failure, and lack transparency. This paper proposes a framework for decentralized access control and auditability of blockchains and tamper resistance to try to make smart contracts work well. Prototype using Ethereum and IPFS: low latency in authorization, better discretion is achieved comparing with the centralized approaches. This paper presents a blockchain technology framework that will facilitate secure data sharing process in cloud-edge learning networks. The framework has smart contracts that enforce the dynamic policy of access to provide auditable records of all data transactions. A prototype implementation that is tested in a simulated federated learning scenario shows how the system can deal with access decisions with low latency and data integrity and audit beyond. These findings indicate that blockchain can be one of the possible ways of moving to decentralized, policybased information cooperation in the education sector, where information protection and institutional trust are critical.