MLWSN 2013 - Special Session on Machine Learning for Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have emerged as an active research area in the last two decades and has already reached considerable maturity. Applications of wireless sensor networks have been developed for a significant number of real-world difficult problems that could be tackled effectively only by the deployment of wireless sensor networks, such as in environmental monitoring (e.g., air quality, water quality, forest or volcanoes monitoring), industrial monitoring and control applications, agriculture, smart home monitoring, health monitoring, military surveillance, to give some examples. Many technical challenges traditionally faced by WSN designers and implementers have been resolved, and the availability of robust sensors and devices as well as adequate communication protocols have made possible the adoption of this technology in the above application areas and many others. The challenge is now not in having effective and cheap sensors and communication, but in the huge deluge of data gathered from sensors, in processing such large amount of data and making sense out of it, given the multiple constraints available in WSNs. Machine Learning (ML) methods have been central in developing WSN applications since the very early days of WSNs, as many of the problems in WSNs could be put as optimization or modeling problems. But, the new task of interpreting, visualizing, and exploiting the big flood of data streamed by WSNs is now emerging as a timely research topic, which provides new opportunities for ML techniques in this area, that will change the traditional perception of WSN systems from being data gatherers only to rather being knowledge generators. This session calls for contributions towards the realization of tomorrow’s WSNs as end-to-end knowledge generation and decision making systems. Research on the theoretical and practical aspects of using ML techniques, such as Bayesian techniques, fuzzy systems, neural networks, evolutionary computing, distributed learning and active learning methods, to WSNs are particularly sought.
TOPICS:
We encourage submission of original research papers on using machine learning methods in wireless sensor network systems, including but not limited to the following topics:
? Use of ML methods for sensor placement, network optimization and other WSN
? optimization related issues
? ML models for embedding predictive capabilities in WSN systems
? Bayesian inference and networks, fuzzy techniques, evolutionary computing/optimization techniques for WSNs
? Information extraction and data mining methods for WSNs
? Practical implementation examples, achievements and challenges in effectively
? deploying in-network data mining and data reduction techniques
? Tools for data interpretation and knowledge generation from WSNs
? Examples of field deployed wireless networked sensing systems with embedded machine learning and decision making models
? Case studies of WSN systems that embed some form of data processing and intelligence to deliver informational and knowledge outputs under real-time or near-real-time constraints