Long Range Wide Area Network For Wireless Underground Sensor Networks

Abstract

Wireless Underground Sensor Network (WUSN) is an active research topic for the past decade. WUSN technologies have a wide range of applications in the field of agricultural and geological monitoring. Traditional sensing network techniques have limited range, lack of concealment, and sometimes require data to be manually retrieved. Recent advancements in Low Power Wide Area Networking (LPWAN) technologies promise long range, low power and lost cost with a compromise of a low data rate. Emerging LPWAN technologies such as Long Range Wide Area Network (LoRaWAN), SigFox and NarrowBand-Internet-of-Things (NB-IoT) all offer potential WUSN solutions where the end devicescan be fully buried and connected with sensors and data can be sent to the gateway wirelessly in real time and then forwarded to a central network server.With LoRaWAN chosen to be the most balanced solution for WUSN amongst the emerging LPWAN technologies, in term of availability, flexibility and performance, LoRaWAN WUSN Simulator (LWS), was developed. To simulate the signal propagation in an underground environment, different soil path loss models were compared, and a modified Friis model was chosen to be implemented in LWS. A field experiment was also conducted to evaluate the accuracy of the theoretical path loss model. LWS is also capable of simulating collision behaviours of LoRa Physical Layer and Medium Access Layer (MAC) layer. The effect of regional duty cycle regulation is also examined in detail to evaluate how it introduces congestion in the network. The simulation results showed that the regional duty cycle restriction significantly reduces the network data extraction rate if the application requires acknowledgements for all or the majority of end-devices. However, for agricultural applications where the end-devices do not require frequent acknowledgements from the network server, the network is still scalable with the duty cycle restriction. Simulation results also showed that a LoRaWAN-based WUSN network is capable of having a network coverage in the multi-kilometres range with a burial depth in the topsoil range of less than 50 cm. This research has demonstrated the potential of LoRaWAN in agriculture WUSN with LWS being a versatile pre-deployment analysis tool.