Techniques for resilient real-world IoT

In this paper, we focus on two methods that can be used to increase the resilience of remotely-deployed and unattended IoT devices that must operate without direct human intervention. The methods include adaptive messaging rates for power-constrained devices where the most energy intensive process involves the transmission and reception of data via wireless means, and data caching and replay for extended network connectivity outages. In this paper, we discuss how these techniques were applied in a real-world wireless IoT sensor network in Dublin, Ireland. Specifically, we show that: 1) The operating lifetime of a power-constrained device can be extended through the use of adaptive messaging rates 2) Sensor data backfilling following extended periods of network outages is a feasible method of preventing sensor data loss. We provide a example demonstrating that the operating lifetime of a power-constrained IoT device can be extended using the adaptive message rate approach. Additionally, we demonstrate how extended network connectivity outages can be recovered from using an elastic message cache approach. Finally, we conclude that the combination of both of these techniques are feasible methods to increase the resilience of real-world IoT devices operating in volatile power and wireless communications environments.