Section: New Results
Routing, deploying in network processing
In [15] xe consider a system composed of a set of mobile sensors, disseminated in a region of interest, which mobility is controlled (as opposed to mobility imposed by the entity on which they are embedded). A routing protocol in this context enables any point of the region to be reached. The strength of our proposition Grasp (GReedy stAteless Routing Protocol), beyond its simplicity, is that routing enables a free and close to optimal self-deployment of sensors over a given region. Grasp transparently copes with dynamic changes of the region of interest. In addition, Grasp is independent from the underlying communication model. Grasp ensures (i) that routing is always possible in a mobile WSN irrespective of the number of sensors and (ii) above a given number of sensors in a considered zone, the protocol eventually ensures that routing does no longer require sensors to move, thus providing self-deployment. In one dimension, Grasp converges to a full connected-coverage of the region with the minimum required number of sensors in a finite number of steps, ensuring an optimal deployment. In two dimensions, sensors reach autonomously a stable full coverage following geometrical patterns. This requires only 1.5 the optimal number of sensors to cover a region. A theoretical analysis of convergence proves these properties in one and two dimensions. Some simulation results matching the analysis are also presented.
In [9] , [18] , we provide a first analysis of the space and time correlations of the collected live E! measurements. We propose an original synchronous average-based model to decompose each time series in order to separate dependencies stemming from expected cyclic trends, such as seasonal (year) and daily effects from dependencies actually observed on residual fluctuations. Then, we investigate the auto- and cross- correlations of both the measurements observed at a given weather station and between measurements collected at adjacent stations.
