Standard of things, first step: Understanding and normalizing sensor signals

Abstract

We live in an interconnected world. The IoT, Internet of Things, is a reality. There are more things connected to internet than people on the planet. The communication channels has been established and standardized, things can be easily identified. We know how to interchange data but we do not know exactly what to interchange. There is a need to reduce the traffic over the networks when the number of nodes increases. Things could be sensors or devices with sensors. Things generate signals and since that the world is inferred through these signals, there is a need to interchange knowledge at low bit rates in a standardized way. The technology explosion that invades our lives has an influence not only on what is plausible to be standardized but opens possibilities not even thought. In the sensor world, intelligence has been concentrated in the point of acquisition, inside the transducer. Considering CPU capabilities, cost, low consumption and reduction of size, the sensor signal will be processed entirely into the smart sensor, even for cheap and simple sensors. The network will be used to interchange more meaningful information or knowledge. This paper focuses on algorithms suitable for any sensor signal that start from oversampled data and end with knowledge about the sensor signal behavior in a normalized scheme. The proposed standard is an important step towards an interconnected and manageable world.