Go to JKU Homepage
The number of wireless communication devices is inexorably increasing to form massive connected networks within the paradigm of the Internet of Things (IoT). IoT devices must have low power consumption and communicate with low latency and low data-rate while enabling self-organizing features during deployment and network operation. These requirements translate to research challenges on the communication protocols, network organization and connectivity. Fundamental aspects for those challenges are time synchronization and geolocation of the devices.
Although synchronization and localization are strongly linked by a packets time-of-flight measured in local time, they are mostly treated independently. Neglecting this connection increases the energy consumption of the nodes and requires additional channel resources for communication. Moreover, in the vision of IoT as vast amount of connected objects, tasks have to be solved in a decentralized way, making the network robust and scaleable. Our research focuses on the unification of the synchronization and localization problem for decentralized networks.
Decentralized Synchronization and Localization in Cooperative Networks
Contact
E-Mail:
Bernhard Etzlinger
Potential Industrial Applications
- Autonomous robot platforms in industrial processes: In a product centric perspective, production processes are services that are offered to the product on a certain location (the production machine) in the plant. The interconnection of those services is solved by logistic units, which offer transportation as a service to the product. To gain a maximum of flexibility and robustness, autonomous robot platforms optimize the logistic task based on their local perception of the environment. A fundamental block is a flexible and robust self-localization method of the robot, paired with learning capabilities and game-theoretic decision finding.
- Self-attribution of sensors in complex measurement tasks: Testing is a major task in product development. A multitude of sensors have to be placed in a small area and correctly attributed to interpret the sensor values they provide. Errors in the attribution are often only discovered at a later state, making collected data sets useless and adding high cost to re-measure the task. When sensors are able to accurately locate themselves, they can autonomously attribute the values they are providing. Due to complex deployment environments and a high density of sensor nodes, low-complex cooperative approaches are a vital solution.
