PROJECT TITLE :

Spatial Field Reconstruction and Sensor Selection in Heterogeneous Sensor Networks With Stochastic Energy Harvesting - 2018

ABSTRACT:

We tend to address the two fundamental issues of spatial field reconstruction and sensor selection in heterogeneous sensor networks. We have a tendency to consider the case where two varieties of sensors are deployed: the first consists of high-priced, prime quality sensors; and therefore the second, of cheap low quality sensors, which are activated solely if the intensity of the spatial field exceeds a pre-outlined activation threshold (e.g., wind sensors). Additionally, these sensors are powered by suggests that of energy harvesting and their time varying energy status impacts on the accuracy of the measurement that will be obtained. We have a tendency to then address the subsequent two vital issues: (i) a way to efficiently perform spatial field reconstruction based on measurements obtained simultaneously from each networks; and (ii) a way to perform query based mostly sensor set selection with predictive MSE performance guarantee. To overcome this problem, we tend to solve the first problem by developing a low complexity algorithm based mostly on the spatial best linear unbiased estimator (S-BLUE). Next, building on the S-BLUE, we have a tendency to address the second downside, and develop an economical algorithm for query primarily based sensor set selection with performance guarantee. Our algorithm is based on the Cross Entropy methodology which solves the combinatorial optimization downside in an efficient manner. We tend to gift a comprehensive study of the performance gain which will be obtained by augmenting the high-quality sensors with low-quality sensors using each artificial and real insurance storm surge database known as the Extreme Wind Storms Catalogue.