Secure monitoring in IoT-based services via fog orchestration

Fog computing proposes moving computation, communication and storage from the cloud to the edge of the network, thus making it a perfect match for latency-sensitive IoT services such as sensor monitoring. Security is, however, a fundamental issue in fog-enabled services dealing with sensitive data, such as healthcare monitoring. Some authors circumvent security issues by assuming that IoT devices and fog nodes are fully trusted. Nevertheless, when IoT devices are geographically distributed and/or may be deployed in non-secure locations, and communications are done via WAN, the probability of attacks increases significantly. Security-enabling solutions rely on either trusted and centralized infrastructures (which go against the distributed and ubiquitous nature of fog-enabled IoT services) or expensive cryptography (which increases latency). In contrast, we propose decentralized and efficient security-enabling protocols for fog-based services that involve continuous monitoring of data. We have put special care in ensuring that the performance benefits brought by fog computing are not significantly hampered by the security mechanisms in use. For this, we rely on the recent concept of fog orchestration by which the network is self-tailored to the service to be delivered; in this way, we minimize the load of the network and its nodes during the monitoring. By means of a set of theoretical analyses and empirical results, we discuss the performance of our protocols and how they improve the current literature.