Identifying the Challenges in Reducing Latency in GSN using Predictors

Andreas Benzing, Klaus Herrmann, Boris Koldehofe, Kurt Rothermel


Simulations based on real-time data continuously gathered from sensor networks all over the world have received growing attention due to the increasing availability of measured data. Furthermore, predictive techniques have been employed in the realm of such networks to reduce communication for energy-efficiency. However, research has focused on the high amounts of data transferred rather than latency requirements posed by the applications. We propose using predictors to supply data with low latency as required for accurate simulations. This paper investigates requirements for a successful combination of these concepts and discusses challenges that arise.

Full Text:




Hosted By Universitätsbibliothek TU Berlin.