Spreading Factor Selection Mechanism for Transmission over LoRa Networks

This paper presents a mechanism for Spreading Factor (SF) prediction in LoRa networks for more optimized data transmissions. The proposed mechanism is based on Machine Learning (ML) algorithms and assigns the node's SF value based on prior transmission data. This paper examines three different approaches for the selection of the SF during LoRa transmissions a) Random SF assignment b) Adaptive Data Rate (ADR) and c) ML based SF selection. The main target is to study and determine the most efficient approach, as well as to investigate the exploitation of ML techniques in the context of LoRa networks. We created a simple library based on ML libraries, such as Scikit Learn that can be used with the FLoRa an OMNeT++ based LoRa simulator. With the use of this library, it is possible to predict the node's SF using ML techniques. Two classification algorithms were tested, the k Nearest Neighbors (k-NN) and Naïve Bayes classifier. Finally, we compared the ML mechanisms with two variants of the ADR mechanism. The approaches performance is evaluated for different scenarios, using the delivery ratio and energy consumption metrics.