Smart offloading technique for CP-ABE encryption schemes in constrained devices

Nowadays, we are attending to the wide proliferation of IoT devices in various computing environments. Actually, this trend is the natural way to extend the M2M communication. However, these devices are facing serious resource constraints. To mitigate this problem, the huge amount of the exchanged data is collected from these devices for being stored and processed by the Cloud. Furthermore, since the Cloud is honest but curious, it may reveal personal information about users’ habits owning these devices and can even lead to user profiling. Consequently, security mechanisms should be deployed at different levels to preserve data privacy so that only authorized users can gain access to the smallest piece of data according to the collection purpose. By assuming that the Cloud-is-honest-but-curious, we can not provide full or similar access to different untrusted Cloud services. Therefore, we should define the relevant privacy level implementing the data access control for each Cloud service according to some criteria. In this context, CP-ABE is a promising solution addressing this problem. This novel attribute-based public key encryption system provides a flexible fine-grained access control to data for any data requestor. However, performing all the related cryptographic operations on such devices is practically infeasible because of the resource constraints. For alleviating all the computation burden on these resource-limited devices, several schemes have been proposed. In this work, we propose a smart offloading technique that switches dynamically from full encryption to partial encryption according to a wise decision strategy considering the available resources and some crucial parameters like the number of attributes and the size of the data being encrypted. The relevant decision is based on a machine learning algorithm. To the best of our knowledge, this is the first paper proposing an adaptive CP-ABE scheme for constrained device optimizing the overall available resources.