Long range low power sensor networks with LoRa sensor for large area fishery environment monitoring

Abstract: The design of wireless monitoring system for fishery environment based on Internet of Things technology has become a research hotspot at present. Aiming at the characteristics of long monitoring period and large coverage area in applications of inshore fishery and large area aquaculture environment monitoring, -the long range low power sensor networks was designed based on LoRa technology-. Due to the expensive price and technology closedness of commercial LoRa gateway SX1301, a low cost open dual-channel gateway was designed, based on two LoRa SX1278 transceivers. Based on the hardware design of dual-channel gateway, a dual-channel synchronous communication scheme with adaptive data rate was proposed. Two LoRa SX1278 transceivers corresponded to two independent channels, namely periodic data channel 1 and special function channel 2. The periodic data channel was used to upload regular packets, data collecting cycle of which was divided into several periods using different spreading factor (SF). Each period was equally split into several slots to the nodes within the same distance range in which the spreading factor was the same. The special function channel 2 was used to upload urgent packets, join in network, and synchronize time. For urgent packets, the terminal nodes could directly upload by switching channel 2 without waiting for the allocated slot arriving , to improve the real-time response of the network system. The scheme fully took the advantage of the characteristics of multiple spreading factors based on LoRa. Different spreading factors was assigned to terminal nodes according to the distances to gateway automatically with the different received signal strength indication and signal noise ratio of the received packets to ensure network connectivity. The scheme was built into a custom Matlab simulator to evaluate the data collision ratio, communication delay and energy cost performance varying with the network size(number of terminal node). With the terminal nodes increasing to 100, the performance of proposed scheme was nearly constant; but the collision ratio with single channel and single spreading factor was nearly linearly increasing to 1, and energy cost increased much more. The field experiment was carried out to test the effectiveness and reliability in Dishui Lake in Shanghai. An application program(App) for Android mobile phone was developed for monitoring field data. Four terminal nodes with dissolved oxygen sensor and pH value sensor were deployed in different distance to gateway. To line-of-sight communication, the average delivery ratio of proposed scheme increased from 0.8 to 0.99 compared with single spreading factor. To the rapid communication with 10 s uploading interval, the average delivery ratio increased from 0.4 to 0.95 compared with single spreading factor of 12. The proposed scheme largely reduced the wireless channel collision probability, improving the real-time performance for urgent data and the average lifetime of terminals, while ensuring large coverage area of monitoring network by the MAC layer synchronization scheduling. The simulation and experiment results showed the effectiveness of the proposed scheme. The designed sensor network system could cover a monitoring region with radius of 3 km, with the capacity of 100 terminal nodes and terminal node with 3600 mA?h lithium battery could work for 1 a with the 10 min data uploading interval.