Secure Analytics and Resilient Inference for the Internet of Things

Abstract

Internet of Things (IoT) applications for Smart Cities, such as systems for traffic control and pollution monitoring, increasingly rely on trustworthy and secure data analytics. Proper countermeasures are needed to ensure that IoT applications function reliably under security threats. This paper studies secure analytics and resilient inference for IoT in the context of recursive parameter estimation. A team of devices makes noisy measurements of an unknown parameter, and an attacker manipulates the measurement data of a subset of the devices. We present a resilient recursive estimation algorithm that processes the measurement streams to recover the value of the parameter, even when a subset of the devices fall under attack. The estimator is guaranteed to be strongly consistent - that is, the estimate converges almost surely to the value of the parameter - as long as less than half of the devices fall under attack. We illustrate the performance of the estimator through numerical examples.