Experimental implementation of a LoRa sensor network and robustness analysis of simulated TX-RX LoRa signals using LabVIEW.

The Internet of Things (IoT) is the network of physical objects that contain embedded technology to communicate and sense or interact with their internal states or the external environment. Low-Power Wide-Area Networks (LPWAN) represent a set of long-range communication technologies suitable for supporting IoT applications, which need multi-year battery lifetime and usually send small amount of data over long distances a few times per hour. LoRa is the first low cost implementation of Chirp Spread Spectrum (CSS) modulation for commercial usage suitable for meeting IoT requirements, such as long range of communication, low power consumption, low data rate, low cost, high level of robustness to interference. Two are the main objectives of this thesis: the former concerns an experimental implementation of a LoRa sensor network installed inside the building of ENSIL (École National Supérieure d’Ingénieurs de Limoges), at the University of Limoges in France; the latter consists in the robustness evaluation of a transmitter-receiver LoRa signal, simulated with the help of LabVIEW. For what concerns the first part, 30 sensors for monitoring temperature and humidity are all wirelessly connected to a gateway using LoRa technology. However, the overall power consumption is too high, therefore the battery lifetime must be increased at the sensor level. The purpose of the infrastructure is thus to convey temperature and humidity information from sensors to a gateway, based on a Raspberry Pi board, which used to upload the data to ThingSpeak cloud. Therefore, after a preliminary analysis of LoRa operating modes, some modifications at the power management level have been applied both on the sensor and gateway side. Moreover, ThingSpeak has been replaced with a new server, based on a MySQL database and a basic web user interface has been created to allow users to access to the database records of interest and eventually to delete them, if they are in charge of it. In the second part, instead, LoRa robustness performance to interference is assessed by means of the bit error rate behaviour under different noise channel conditions. In order to realize a LabVIEW simulation of LoRa physical layer without using real hardware transceivers, a deep analysis of the literature was essential. Indeed, due to its proprietary license nature, several studies tried to reveal more details about LoRa PHYsical layer (PHY) layer, intercepting on-the-air LoRa signals by means of a Software Defined Radio (SDR) enabled LoRa gateway. Their results and conclusions have been exploited to implement the TX-RX LoRa signal, with some simplifying assumptions on the frame synchronization and the channel coding. Finally, the study has been conducted considering the Bit Error Rate (BER) performance with different parameter conditions and variations. The impact of two types of noise source was taken into account, i.e. Additive White Gaussian Noise (AWGN) and narrowband interference.