Integrating Sensor Streams in pHealth Networks

Personal Health (pHealth) sensor networks are generally used to monitor the well being of both athletes and the general public to inform health specialists of future and often serious ailments. The problem facing these domain experts is the scale and quality of data they must search in order to extract meaningful results. By using peer-to-peer sensor architectures and a mechanism for reducing the search space, we can, to some extent, address the scalability issue. However, synchronisation and normalisation of distributed sensor streams remains a problem in many networks. In the case of pHealth sensor networks, it is crucial for experts to align multiple sensor readings before query or data mining activities can take place. This paper presents a system for clustering and synchronising sensor streams in preparation for user queries.