Related PhD projects

Lukasz Sergiusz Michalik: 

"Connectivity of Edge-Devices in Resource-limited Environments"

The project investigates different connectivity methods for resource-limited devices; called Observation Units; observing and monitoring the Arctic Tundra. There is a need for a connectivity system that will improve the process of data collection from remote devices deployed in the environment that is accessible only for a short period of time over the year and where energy harvesting is limited. Cellular coverage in these regions either does not exist or is highly limited. This requires researchers to build new "communication infrastructure" that does not integrate into the natural environment. Many monitored sites are located in national parks where rules and regulations restrict any human constructions. This means that the infrastructure should be integrated into Observation Units themselves. To provide flexibility for multiple applications these devices could be equipped in a multi-radio capabilities controlled by the connectivity system developed to introduce wireless communication between edge-devices and data stores. The system could provide autonomy for observation devices by being aware of an available energy, different radios characteristic and different ways the radios could be utilized. This project is funded by the departement of computer science.

Mike Murphy:

"Consistency AbstractionPh.D-fellow s for Distributed Data Over Delayed and Disrupted Networks"

Mike is focusing on the nuances of gathering observation data over difficult network conditions. Networking in the tundra will involve constrained and unreliable connections, and may require relays, multiple routes, and even manual hops over a skier-net to get to a back-haul network. How do we keep observation data consistent as it slowly propagates through such a network? And how do we make such a solution bandwidth and energy efficient for use on constrained devices? How do we keep users apprised of what new data has arrived, what data is on the way, and what data needs manual intervention to retrieve? How do we conceptualize our observation data set as a cohesive whole, even though it is scattered across the tundra?

Mike is also the group member who is working most closely with COAT's ecologists, joining them in the field to see firsthand what it takes to collect ecological data in the tundra. As an exploratory project to gain real-world tundra experience, Mike and master's student Øystein Tveito have designed and built simplified Observation Units to measure temperature and CO2 levels in lemming habitats under the snow. Mike and Øystein have deployed ten of these CO2 Units in the low-arctic tundra of Norway's Varanger Peninsula, where five of them are picking up just enough LTE CAT-M1 signal to send data home. The others will have to wait for next summer's fieldwork hiker-net and further progress on our DAO connectivity, consistency, and update solutions. This project is funded by COAT tools project.

Roberth Tollefsen:

Roberth's PhD project will research and develop a system for automatic updates of groups of observation units (OU). The goal is for the OU to be autonomous and handle updating of configurations and software. To handle updates the OU may need to schedule updates, handle failures both in transfer, the update itself (tampering) and bugs/crashes in the software. This project is funded by the DAO project.

Steffen Ole Randrup Kristensen:

To come ...
This project is funded by the DAO project.