Implementation of an Age of Information Aware Wireless Sensor Networks
Open Access
- Author:
- Chache, Frederick
- Graduate Program:
- Electrical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- December 01, 2023
- Committee Members:
- Madhavan Swaminathan, Program Head/Chair
Karl Reichard, Outside Unit & Field Member
Ram Narayanan, Chair & Dissertation Advisor
Daniel Cullina, Major Field Member
Mahanth Gowda, Major Field Member
Ramesh Bharadwaj, Special Member - Keywords:
- Age of Information
Wireless Sensor Networks
LoRa
Internet of Things - Abstract:
- Wireless internet connected sensor networks have a wide variety of applications in sensing and communications. The range of applications is generally limited by a device's ability to have and maintain internet access, which constrains devices to more indoor and urban environments. Advances in 5G technology and in low power RF modulation have provided the tools to greatly expand the range of environments in which Internet of Things (IoT) devices can operate. For example, the agriculture industry increasingly demands IoT devices to enable automation, but these typically operate in environments with minimal digital infrastructure, where internet coverage is either nonexistent or cost prohibitive for low cost sensors. This dissertation demonstrates how a range of new and established technologies can be leveraged to develop a low cost, low power, and resilient network for digital applications in rural environments. Long Range (LoRa) is a low power, long range radio frequency (RF) modulation scheme with low data rates which provides the physical layer backbone of the architecture. Fifth generation (5G) mobile network channel encoding technologies, such as channel encoding and data compression, are leveraged to improve data rates and effective ranges. Furthermore, mesh networking technologies can be used to allow nodes to generate independent networks in real time, acting as gateway nodes to share internet connection with more distant nodes. These technologies all possess inherent drawbacks that must be balanced in order to develop a network capable of delivering data that are useful and timely to the back end applications requiring the sensor data. To this end, the Age of Information (AoI) paradigm was utilized to provide a performance metric for these applications. Extensive work was done to extend the existing AoI models to encompass the new technologies. Using theses approaches, extensive simulation and theoretical models could be developed to study the effects of the new technologies and enable the development of more refined algorithms. In this work, an architecture was developed that utilizes these technologies to prove its effectiveness. This architecture was tested using low cost hardware to validate the theoretical models using the measured results.