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

Abstract

We address the two fundamental problems of spatial field reconstruction and sensor selection in heterogeneous sensor networks. We consider the case where two types of sensors are deployed: the first consists of expensive, high quality sensors; and the second, of cheap low quality sensors, which are activated only if the intensity of the spatial field exceeds a pre-defined activation threshold (eg. wind sensors). In addition, these sensors are powered by means of energy harvesting and their time varying energy status impacts on the accuracy of the measurement that may be obtained. We account for this phenomenon by encoding the energy harvesting process into the second moment properties of the additive noise, resulting in a spatial heteroscedastic process. We then address the following two important problems: (i) how to efficiently perform spatial field reconstruction based on measurements obtained simultaneously from both networks; and (ii) how to perform query based sensor set selection with predictive MSE performance guarantee. We first show that the resulting predictive posterior distribution, which is key in fusing such disparate observations, involves solving intractable integrals. To overcome this problem, we solve the first problem by developing a low complexity algorithm based on the spatial best linear unbiased estimator (S-BLUE). Next, building on the S-BLUE, we address the second problem, and develop an efficient algorithm for query based sensor set selection with performance guarantee. Our algorithm is based on the Cross Entropy method which solves the combinatorial optimization problem in an efficient manner. We present a comprehensive study of the performance gain that can be obtained by augmenting the high-quality sensors with low-quality sensors using both synthetic and real insurance storm surge database known as the Extreme Wind Storms Catalogue.

Simultaneous View and Feature Selection for Collaborative Multi-Robot Perception

“…For target tracking, sensors can be chosen based on their capabilities of location [38], [39] or to best provide coverage of an entire area [40]. Selection methods have also been introduced based on task allocation algorithms [41], sparsity [8], [42], cross entropy [43], and even randomized algorithms [44], [45].…”

Joint Sparse Observation and Coding Design for Multiple Phenomena Monitoring

“…Considering the main energy consumption of IoT, the work of [10] studied the optimal node selection for data transmission with energy conservation. Utilizing energy harvesting as the power supply, the node selection approach was adopted in [11] for heterogeneous sensor networks to achieve energy-efficient spatial field reconstruction. To improve energy efficiency, grouping the selected nodes into multiple clusters is an effective method [12]- [15], which can maintain a longer life of distributed nodes by reducing the power loss in propagation and retransmission.…”

Dynamic Weight Importance Sampling for Low Cost Spatiotemporal Sensing

“…Energy conservation is a challenging problem in wireless sensor networks [7]. Sensor selection in sensor network for spatial signal modeling [8] is an effective way to conserves the in-network energy. Modeling the spatial signal using contour lines is an example for energy conservative sensor selection.…”