Convex Optimization and Model Identification for Reliable and Energy Efficient QoS-Aware IoT Systems

In 2020 there will be more than 50 billions of interconnected devices composing the Internet of Things (IoT) to enable fine-grained sensing and actuation in a variety of fields, from smart environments to healthcare. A large part of IoT devices are powered by batteries that need to be changed or recharged frequently. Also, IoT networks require replacement of defective parts for meeting Quality of Service (QoS) targets, with high associated costs. Maintenance and energy efficiency represent important limiting factors to such growth. In this paper we formulate a comprehensive problem for reducing IoT energy consumption subject to QoS and reliability constraints. We then propose a framework to dynamically manage reliability and energy consumption using convex optimization. The proposed framework leverages model identification techniques for efficiently identifying reference values at runtime.