将WSN和 machine learning 相结合进行研究
wireless ad hoc networking (sensor networks, vehicular networks, automotive networks, robotic networks, etc.) and the application of machine learning and computational intelligence techniques to various communication and application level problems. I am also highly interested and actively researching in the area of wireless sensor networks deployment and performance analysis, including testbeds and simulation techniques.
Publications
Book chapters
Anna Foerster, Amy L. Murphy: Machine Learning Across the WSN Layers, in Wireless Sensor Networks, book chapter in Wireless Sensor Networks, ISBN 978-3-902613-49-3,InTechWeb Publishing, editors A. Foerster and A. Foerster, to appear in eary 2010.
Conference proceedings
Anna Foerster, Alexander Foerster, Amy L. Murphy: Optimal cluster sizes for wireless sensor networks: An experimental analysis, Proceedings of the 1st International Conference on Ad Hoc Networks, Niagara Falls, Canada, 2009. to appear.copy upon request
Anna Foerster, Amy L. Murphy: Clique: Role-Free Clustering with Q-Learning for WSNs, Proceedings of the 29th International Conference on Distributed Computing Systems (ICDCS), Montreal, Canada, June 2009.
PDF (456 KB)
Anna Foerster, Amy L. Murphy, Jochen Schiller, Kirsten Terfloth: An Efficient Implementation of Reinforcement Learning Based Routing on Real WSN Hardware, Proceedings of the 4th IEEE International Conference on Wireless and Mobile Computing, Networking and Communications, Avignon, France, October 2008.
PDF (200 KB)
Technical reports
Anna Foerster, Amy L. Murphy: FROMS: A Failure Tolerant and Mobility Enabled Multicast Routing Paradigm with Reinforcement Learning for WSNs, Technical Report of the University of Lugano 2009/04, June 2009.PDF (1.6 MB)
Anna Egorova-Foerster, Amy Murphy: A Feedback Enhanced Learning Approach for Routing in WSN, Technical Report of the University of Lugano 2006/03, May 2006.
PDF (350 KB)
Thesis
Anna Foerster: Teaching Networks How To Learn - Data Dissemination for Wireless Sensor Networks with Reinforcement Learning, PhD thesis at the University of Lugano, Switzerland, May 2009.
copy upon request
Protocols and add-ons for MF
- LMAC, BMAC and a battery model for WSNs by Anna Förster. More info and download:http://www.inf.unisi.ch/phd/foerster/downloads.html
- FeedbackFramework - a feedback-enhanced simulation environment by Anna Egorova-Förster. This is not a standard routing protocol, however, if configured properly, it could be also a directed diffusion "one-phase pull" implementation. More info and download: http://www.inf.unisi.ch/phd/foerster/downloads.html
- Feedback-enhanced routing protocol implementation by Anna Egorova-Förster. It is a routing/application framework over Omnet++ and the MF (the newest preview), which enables fast prototyping and implementataion of different routing strategies, which use feedback exchange. It includes a Directed Duffusion's one-phase-pull protocol implementation. More info and download: http://www.inf.unisi.ch/phd/foerster/downloads.html
- Creating random connected topologies. MatLab function implementation by Anna Egorova-Förster to produce many random connected topologies for the Mobility Framework. It calculates the interference distance just as the ChannelControl does, so the connectivity is guaranteed (more or less..) The difference to other methods presented so far is that it produces a completely random topology and then checks whether it is connected or not (simple broadcast to all nodes). You'll need of course a MatLab installation to run it. Download from: http://www.inf.unisi.ch/projects/mics/downloads.html
