Adaptive data update for cloud-based internet of things applications

In the Internet of Things (IoT) new challenges for both the networks and the applications are arising. This is due to the use of constrained communication platforms, like Low Power Wide Area Networks (LPWANs), due to the handling of variable and heterogeneous data traffic and due to the fact that IoT applications are characterized by sporadic interactions of IoT devices. Above all, providing simultaneous connectivity among billions of devices will revolutionize the concept of data exchange and sharing. A fundamental goal in this context is to limit the IoT power consumption and reduce the traffic load that can be generated by many devices. This task can be implemented at different protocol layers. By referring to the application layer, in this paper we present an adaptive data update algorithm, named ED-rsync, aimed at reducing the information volume exchanged between IoT devices, transmitting some data/measurements, and the application server hosted in the cloud. The benefit brought by the proposed solution is twofold. On the one hand, we reduce the uplink traffic, that is the flow from the IoT device to the cloud service. On the other hand we avoid to use the downlink that typically in IoT networks is expensive and may be also very inefficient. Traffic optimization entails also device power saving that represents one of the key challenges in the IoT context. So, following this direction, we analyze the impact of ED-rsync on the IoT device energy consumption, reporting some numerical results related to scenario considering three main LPWAN technologies, namely Sigfox, LoRa and Narrowband-IoT.