Marcus Chang har opnået ph.d.-graden i datalogi – Københavns Universitet

Datalogisk Institut, DIKU > Nyheder > DIKU-nyheder 2010 > Marcus Chang har opnåe...

29. januar 2010

Marcus Chang har opnået ph.d.-graden i datalogi

Marcus Chang forsvarede fredag den 29. januar 2010 sin ph.d.-afhandling med titlen  

From Automatic to Adaptive Data Acquisition
- towards scientific sensornets

Bedømmelseskomiteen, som bestod af:

Professor Brian Vinter, DIKU
Professor: Gabor Karsai (Vanderbilt University, USA) samt

Professor: Gaurav S. Sukhatme (University of Southern California, USA)

var enige om, at Marcus Chang havde ydet en flot præstation og fortjente sin ph.d.-titel. Se abstract nedenfor.


Sensornets (network of wireless sensors) have been used for ecological monitoring the past decade, yet the main driving force behind these deployments are still computer scientists. The denser sampling and added modalities offered by sen-sornets could drive these fields in new directions, but not until the domain scientists become familiar with sensornets and use them as any other instrument in their toolbox.

We explore three different directions in which sensornets can become easier to deploy, collect data of higher quality, and offer more flexibility, and we postulate that sensornets should be instruments for domain scientists.

First, as a tool to ease designing and deploying sensornets, we developed a methodology to predict applications' resource consumption on different hardware platforms, without actually having to execute them.

Second, in order to reduce the amount of faulty and missing measurements, we developed an anomaly detection frame-work based on machine learning. This allows faulty measurements to be detected immediately while the ecological ex-periment is running.

Third, to increase the flexibility of sensornets and reduce the complexity for the domain scientist, we developed an AI-based controller to act as a proxy between the scientist and sensornet. This controller is driven by the scientist's require-ments to the collected data, and uses adaptive sampling in order to reach these goals.

Academic supervisor:

Philippe Bonnet, IT-University

For an electronic copy of the thesis, please contact Dina Riis Johannesen, 35 32 14 23,