Privacy-preserving data search with fine-grained dynamic search right management in fog-assisted Internet of Things

Abstract

Fog computing, as an assisted method for cloud computing, collects Internet of Things (IoT) data to multiple fog nodes on the edge of IoT and outsources them to the cloud for data search, and it reduces the computation cost on IoT nodes and provides fine-grained search right management. However, to provide privacy-preserving IoT data search, the existing searchable encryptions are very inefficient as the computation cost is too high for the resource-constrained IoT ends. Moreover, to provide dynamic search right management, the users need to be online all the time in the existing schemes, which is impractical. In this paper, we first present a new fog-assisted privacy-preserving IoT data search framework, where the data from each IoT device is collected by a fog node, stored in a determined document and outsourced to the cloud, the users search the data through the fog nodes, and the fine-grained search right management is maintained at document level. Under this framework, two searchable encryption schemes are proposed, i.e., Credible Fog Nodes assisted Searchable Encryption (CFN-SE) and Semi-trusted Fog Nodes assisted Searchable Encryption (STFN-SE). In CFN-SE scheme, the indexes and trapdoors are generated by the fog nodes, which greatly reduce the computation costs at the IoT devices and user ends, and fog nodes are used to support offline users’ key update. In STFN-SE scheme, the semi-trusted fog nodes are used to provide storage of encrypted key update information to assist offline users’ search right update. In both schemes, no re-encryption of the keywords is needed in search right updates. The performance evaluations of our schemes demonstrate the feasibility and high efficiency of our system.