Big Data Driven Secure IoT Analytics with Trusted Execution Environments

Abstract

The growing adoption of IoT devices in our daily life engendered a need for secure systems to safely store or analyze sensitive data, as well as a decentralized data processing system to handle vast amount of streaming data. The cloud services used to store data and process sensitive data are often come out to be vulnerable to outside threats. Moreover, to analyze enormous streaming data swiftly, they are in need of a fast and efficient system. In this paper we propose a framework to maintain confidentiality and integrity of IoT data, which is of paramount importance, and manage large-scale data anaytics. We design the framework to preserve data privacy utilizing Trusted Execution Environment (TEE) such as Intel SGX, and end-to-end data encryption mechanism. In addition, we utilize Apache Spark for fast real-time streaming data processing from many IoT devices. We evaluate the framework by performing simple decision making in the SGX securely that involves multiple IoT devices, and a real-time anomaly detection in the streaming data from IoT devices using Spark.