The accelerating growth of the Internet of Things (IoT) has led to the development of many different communication protocols to enable the most optimal environment for the nature of IoT. Requiring long range and low power communication abilities, has resulted in LPWANs (low-power WANs); one such LPWAN is LoRaWAN (Long Range WAN) which is studied in depth in this thesis.The thesis is done in collaboration with The Things Network (TTN), a crowd-sourced LoRaWAN expansing over 6 continents. The data retrieved from the TTN NOC (Network Operations Centre) has played a crucial role in this thesis, as it provides the basis for studying the working and performance of a LoRaWAN.The aim of this study is to use performance evaluation of the network to develop an Adaptive Data Rate (ADR) model which will modify the transmission parameters to improve the performance of the network, in terms of the ratio of received and sent packets (called Data Extraction Rate - DER), while attempting to maintain minimal cost of transmission, in terms of transmission power and usage of bandwidth by measuring the airtime used in transmission.Extensive evaluation and analysis of the NOC data is performed and detailed in this report, followed by a modified ADR model which theoretically will improve the DER of the network. This ADR model is verified by modelling the performance of devices on the network with the usage of the proposed ADR model and without. This theoretical verification proves that the DER of the network improves when the transmission parameters are varied in accordance with the proposed ADR model.