University of California, Riverside (Student Affairs)
High performances wireless embedded systems are a natural choice for monitoring applications which generate and analyze large amounts of data. Such applications include monitoring of large structures, habitats and for surveillance. Getting consistent results when monitoring phenomena is a challenging but critical task for solar powered wireless high power embedded systems. The authors' algorithm relies on an energy predictor to achieve uniform monitoring over time while maximizing the number of tasks executed. Their approach outperforms state of the art algorithms by increasing the number of daily measurement by 30% and reducing their standard deviation by 5.5x.