Search

Wireless cognitive sensor networks arranged in smart community intelligently use the idle channels that alleviates spectrum scarcity. To enhance the connectivity of network, the existing spectrum-aware clustering algorithms group sensors into clusters due to the sensed idle channels of each sensor. However, a clustering metric should depend on various evaluation criteria. This poster proposes a novel spectrum-aware clustering algorithm to obtain more reasonable clustering, where a new weighted clustering metric is used to elect clusterheads. After clustering, the clusterheads sense spectrum instead of all nodes that greatly reduces the energy consumption and efficiently increase the network throughout.

The grid-based target localization for internet of vehicle has the property of sparsity, which can be transformed to a sparse recovery problem due to compressive sensing. The accurate target localization relies on the sufficient measurements, but we cannot determine the number of measurements without knowing the number of targets. In this poster, we propose a novel target localization algorithm based on sequential compressed sensing to select the optimal number of measurements and estimate target locations via l p -norm (0 0 -norm or l 1 -norm minimization.