Fair allocation for transmission parameters to achieve scalability in LoRaWAN

LoRa and LoRaWAN are promising solutions for the upcoming challenges that the Internet of Things (IoT) presents. But even with the recent developments in LoRaWAN, there are still problems to overcome due to the scale of applications required for IoTs. LoRa and LoRaWAN provide many of the desired characteristics for IoT such as long-range transmissions, low power use, and low device cost, but also have an issue with node capacity. Scaling LoRaWAN networks for the capacity that the IoTs desires is the current challenge, so therefore, I aim to improve current solutions to meet this challenge. Since LoRaWAN is a new technology, research in the area is in early development, with some contributions existing but no complete solution that can be applied to a diverse range of LoRaWAN applications. My research extended upon previous research focused on the fairness of collision probabilities across nodes in a network, and thus, I created a solution that improved the scalability of LoRaWAN. Overall, I contributed a parameter allocation algorithm, a transmission power assignment scheme, and a path loss rule for node placement for LoRaWAN networks that, in simulation, achieved better performance than current solutions. This was possible due to lack of consideration regarding the importance of LoRa spreading factor parameters by current solutions and the realistic implications of path loss. By factoring for these two aspects, a more adaptable solution can be created. This solution can effectively maximise LoRaWAN's most important capability and also consider the physical limitations of the real-world like path loss. The proposed solution, Fair Allocation for Transmission Parameters (FATP) for LoRaWAN, adapts the idea of fair device performance in a network to the realistic limitations a real-world application would introduce, such as node distribution and path loss. Firstly, the parameter allocation algorithm is responsible for understanding the path loss of every device and the ideal method of leveraging the spreading factor parameter to achieve a fair network. A fair network, in this instance, is one where every device has the same probability for collision when transmitting. This algorithm then adapts the ideal method of assigning the spreading factor parameter to devices based on their path loss so that the network remains as near to optimal as possible, while respecting the limitations of the real world. Secondly, the transmission power scheme is responsible for minimising the power usage of devices by controlling how much power is used to transmit with. This power scheme functions by minimising every device's transmission power to the minimum amount that will still guarantee a viable transmission. Finally, the path loss rule is a guideline to follow when placing devices in a network. This guideline describes the minimum path loss a device should be allowed to have before it becomes impossible to mitigate its interference with other devices. In the simulation, FATP achieved better performance in terms of data extraction rate (DER) and reducing the variance of performance across nodes in a network for a range of node distributions when compared with current solutions. FATP provided up to a 9% improvement for network DER with an average of a 4% network DER improvement when compared to the nearest best performing current solution, an RSSI-based solution. Most notably, FATP, on average, reduced the variance of individual node performance by a factor of approximately 59. FATP achieved this while consuming approximately the same amount of energy as the RSSI-based solution. Overall, FATP achieved its goals of improving network DER to increase scalability while also ensuring that nodes operate more fairly with more consistent DER performance