Secure Shared Continuous Query Processing

Abstract

Data Stream Management Systems (DSMSs) are being used in diverse application domains (e.g., stock trading), however, the need for processing data securely is becoming critical to several stream applications (e.g., patient monitoring). In this paper, we introduce a novel three stage (pre-processing, query processing, and post-processing) framework to enforce access control in DSMSs. As opposed to existing systems, our framework allows continuous queries to be shared when they have same or different privileges, does not modify the query plans, introduces no new security operators, and checks a tuple only once irrespective of the number of active continuous queries. In addition, it does not affect the DSMS quality of service improvement mechanisms as query plans are not modified. We discuss the prototype implementation using the MavStream Data Stream Management System. Finally, we discuss experimental evaluations to demonstrate the low overhead and feasibility of our approach.