LoRa Channel Propagation Modelling using Artificial Neural Network

In Long Range (LoRa) wireless communication, the transmitted signals may experience loss due to many obstacles as well as interferences. The existing path loss propagation models which used for applications as the one used for LoRa networks are designed to make it simple for calculation and thus making it less accurate. Enabling technology such as artificial intelligence is capable to process complex variables in a very short duration with almost accurate prediction. By having enough data or collecting more data over certain period, this technology can be implemented to learn all the data rules and behavior to predict the output. In this paper, the Artificial Neural network model is proposed to predict the propagation loss of LoRa communication link. Results are compared against path loss propagation models and RMSE values are also determined. The proposed model shows improvement compared with other models in terms of Received Signal Strength Indicator (RSSI) performance and RMSE values.