Use of Unsupervised Learning Clustering Algorithm to Reduce Collisions and Delay within LoRa System for Dense Applications

Abstract

Internet of Things (IoT) is one of the most cited terms within the wireless communication research communities. Next generation wireless networks technologies are expected to have massive-connections of tens of billions of devices. In terms of wireless networks, and in regards to collisions and transmission delay drawbacks being critical challenges when deploying IoT devices, Low Power Wide Area Networks (LPWAN) technologies are considered to be a potential solution for IoT applications. In particular, this paper investigates the use of Long-Range (LoRa) technology for serving dense applications. Furthermore, it identifies a dense application and investigates the possibility of using LoRaWAN for such applications. This work proposes a priority scheduling technique based on unsupervised learning clustering algorithm (K-Means). The proposed technique shows a reduction of the collision rate, the transmission delay and enhancement of the throughput in comparison to conventional LoRaWAN networks and other optimisation techniques.