Work

Large Scale Analysis and Scalability Enhancements for Low Power Wide Area Networks

Year: 2020
Type: dissertation
Abstract: The developing Internet of Things is leading to a much broader range of connected\ndevices across many different industry sectors. The new paradigm of Low Power Wide\nArea Networks (LPWAN) contributes to this by enabling the deployment of cheap lowpower\nwireless devices, allowing for pervasive smart city applications. However, research\ninto LPWANs is in the preliminary stages, with much analysis to be done to quantify\nthe actual limits of proposed protocols. The energy efficiency and scalability of these\nprotocols are key if these technologies are to be deployed effectively in densely populated\ncities. The goal of this work is to quantify and improve the LPWAN paradigm in these\ntwo dimensions.\nThe first contribution is an energy consumption analysis of the LoRaWAN protocol. A\nstudy of the comparative energy consumption rate of the primary LPWAN technologies\nis performed, based on the physical and MAC layer states of the protocols. Then, an\nanalysis of the feasibility of the powering of LoRaWAN applications fully through the\nharvesting of ambient RF energy is presented.\nThe next contribution is a quantification of the energy efficiency and scalability of bidirectional\nLPWANs, using LoRaWAN as a case study. This is performed through the\nimplementation of a LoRaWAN energy model, LoRaWAN Class B, and the Adaptive\nData Rate (ADR) of LoRaWAN in simulation. This work for the first time fully implements\npreviously unsimulated sections of the protocol, enabling the simulation of\nrealistic large scale LoRaWAN networks. Simulations of LoRaWAN networks under a\nvariety of conditions are performed, with large networks of devices running applications\nwith realistic traffic requirements. This analysis can be considered state-of-the-art as it\nconstitutes the first time the analysed LoRaWAN features have been implemented in the\nused simulator.\nThrough the previous analysis, bottlenecks to performance for LoRaWAN networks are\nidentified. The next contribution consists of enhancements to the ADR of LoRaWAN,\nwhich increases the scalability of the network and fixes issues which currently can limit\nthe network adaptability. The final contribution is a novel lightweight collision avoidance\nalgorithm which mitigates the effects of collisions from predictive traffic, based on the\nproperties of LPWAN traffic. The scheme increases the scalability of the network while\nmaintaining the energy efficiency of the protocol.\nThe work presented in this thesis enhances the sustainable operation of LPWANs in\ndenser environments. This work is validated through analytical models and simulation.\nThe applicability of this work to other network wireless protocols is examined and potential\nfuture research directions are discussed. The contributions presented in this thesis\nadvance the evolution of LPWANs by enabling the analysis and enhancing the scalability\nof large scale deployments, further accelerating the acceptability and pervasiveness\nof LPWAN networks.
Language: English
Open Access status: green
OpenAlex ID: W7001593543
Cites: 299
Cited by:
Related to: 9