Selection of Sensors Number Based on Localization Accuracy in Wireless Sensor Networks

In the application of wireless sensor network (WSN), it is of great significance to minimize the number of sensors without deteriorating the localization accuracy. In this article, the number of sensors is identified by the localization accuracy of a single or multiple objects. Bayesian compressed sensing (BCS) is employed as the localization method, which is motivated by the advantage of BCS such as better performance of reconstruction in noisy case. Based on the spatial sparsity of objects to be located in a WSN (comparing with the number of sensors), the localization problem can be converted into recovering a spare index vector using Bayesian estimation. The proposed method can achieve the localization performance without the prior information of objects number. Besides, the minimal level of the sensor number is obtained which can satisfy the location accuracy. The simulation results show that the number of the sensors can be quantified according to localization accuracy in different size of regions and different number of objects.