A Parallel Secure Flow Control Framework for Private Data Sharing in Mobile Edge Cloud

Abstract

Nowadays, the rapid development of edge computing is accelerating the data sharing between cloud computing platforms and mobile users. These data often contain sensitive information, which faces severe leakage risks not only from the semi-trusted cloud servers but also from the malicious senders in the organizations. Fortunately, access control encryption (ACE) has been utilized to secure the data with access control policies, in which a sanitizer (e.g., the edge node) is employed to check all the communications between the sender and receiver, and drop illegal ciphertexts according to the access control policy. However, previous schemes have some limitations in mobile edge cloud, e.g., the sender's attributes are not strictly authenticated in the attribute-based access control policy, or the sanitization time is the bottleneck of fast data sharing. To this end, we introduce PSFlow, a parallel secure flow control framework for private data sharing in mobile edge cloud. First, we propose an attribute-based outsourced ACE (AOACE) scheme, which achieves secure fine-grained data read and write control, and reduces the computational cost of the sender and receiver with outsourced computations in edge nodes. Then, we propose a concrete construction of PSFlow from AOACE, and accelerate the sanitization process with parallel computing. Specifically, PSFlow parallelizes the sanitization operations with a multi-server model in each edge node, and optimizes the sanitization efficiency in each edge server by constructing a shared pool from the attribute universe. The experimental results show that PSFlow is more efficient and practical than previous schemes in mobile edge cloud.