Andreina Liendo Sanchez
PhD, October 2018
Tuesday 11 June 2019
The Internet of Things (IoT) is announced as the next big technological rev- olution where billions of devices will interconnect using Internet technologies and let users interact with the physical world, allowing Smart Home, Smart Cities, smart everything. Wireless Sensor Network are crucial for turning the vision of IoT into a reality, but for this to come true, many of these devices need to be autonomous in energy. Hence, one major challenge is to provide multi-year lifetime while powered on batteries or using harvested energy. Bluetooth Low Energy has shown higher energy efficiency and robustness than other well known WSN protocols, making it a strong candidate for implementation in IoT scenarios. Additionally, BLE is present in almost every smartphone, turning it into perfect ubiquitous remote control for smart homes, buildings or cities. Nevertheless, BLE performance improvement for typical IoT use cases, where battery lifetime should reach many years, is still necessary.
In this work we evaluated Bluetooth Low Energy performance in terms of latency and energy consumption based on analytical models in order to optimize its performance and obtain its maximum level of energy efficiency without modification of the specification in a first place. For this purpose, we proposed a scenarios classification as well as modes of operation for each scenario. Energy efficiency is achieved for each mode of operation by optimizing the parameters that are assigned to the Bluetooth Low Energy nodes during the neighbor discovery phase. This optimization of the parameters was made based on an energy model extracted from the state of the art. The model, in turn, has been optimized to obtain latency and energy consumption regardless of the behavior of the nodes at different levels: application and communication. Since a node can be the central device at one level, while it can be the peripheral device at the other level at the same time, which affects the final performance of the nodes.
In addition, a novel battery lifetime estimation model was presented to show the actual impact that energy consumption optimization have on nodes lifetime in a fast (in terms of simulation time) and realistic way (by taking into account empirical data). Performance results were obtained in our Matlab based simulator based on OOP paradigm, through the use of several IoT test cases. In addition, the latency model used for our investigation was experimentally validated as well as the proposed parameter optimization, showing a high accuracy.
After obtaining the best performance possible of Bluetooth Low Energy without modification of the specification, we evaluated the protocol performance when implementing the concept of Wake-Up radio, which is an ultra low power receiver in charge on sensing the communication channel, waiting for a signal addressed to the node and then wake the main radio up. Thus, the main radio which consumes higher energy, can remain in sleep mode for long periods of time and switch to an active mode only for packet reception, therefore saving considerable amount of energy. We demonstrated that Bluetooth Low Energy lifetime can be significantly increased by implementing a Wake-Up radio and we propose a modification of the protocol in order to render this protocol compatible with an operating mode which includes a Wake-Up radio. For this, we studied the Wake-Up radio state of the art and evaluated Bluetooth Low Energy devices lifetime when a selected Wake-Up radio is implemented at the master side.