An Efficient Supply–Demand-Aligned and Trustworthy MultiKeyword Search Scheme in Edge-Assisted IoT Environments

Abstract

With the integrating development of Internet of Things (IoT) and edge computing, data sharing among various IoT devices has become the trend for extensive applications. However, data sharing in IoT environments is challenged by limited terminal resources and distributed data storage, which places higher demands on security and effectiveness. Even though existing searchable encryption technologies provide feasible solutions, there remain challenges in terms of trustworthy retrieval and execution efficiency. To address these issues, this paper proposes an efficient supply-demand-aligned and trustworthy multi-keyword (ESTM) search scheme in edge-assisted IoT environments, where encrypted documents are stored in edge servers. Furthermore, blockchain-based smart contracts are employed so that search results are consensus on the Fabric ledger and data users can verify whether the returned encrypted documents are reliable using encrypted hashes. To achieve the supply-demand-aligned requirement, the RoBERTa (Robustly optimized BERT approach) model is introduced for text classification and data users can judge which edge server stores data best suits their demands. Meanwhile, coordinate (COO) format is adopted into index vectors and search vectors, which can decrease the time required for constructing an index tree to about 2.7% and the time required for generating trapdoors to about 4.6%. Finally, we conducted an in-depth security analysis and performance comparison with existing works, results show that the proposed scheme is effective and feasible.