Luigi Bruno, Stefano Marano, and Vincenzo Matta
Tracking in Wireless Sensor Networks, Fisher Information, Variable Sampling Rate, Adaptive Signal Processing
The tracking of a target is performed by a multitude of sensors, randomly deployed over a surveyed area, that deliver the measured data to a fusion center. In the estimation step of the tracking procedure, the sensors exploit the a-priori information about the target in order to optimize their measurement frequencies. This is done by maximizing the information provided by each sample, with an energy constraint related to the total number of samples collected by the system. A system design is proposed in which the sensors adaptively tune their measurement rates, resulting in a variable sampling rate system. We formulate two different optimization problems for the sampling rate and derive the solutions in terms of informational quantities related to the Fisher matrix. Computer experiments are presented to corroborate the proposed approach.
Important Links:
Go Back
