An Efficient Compartmental Model for Real-Time Node Tracking Over Cognitive Wireless Sensor Networks

In this paper, an efficient compartmental model for real-time node tracking over cognitive wireless sensor networks is proposed. The compartmental model is developed in a multi-sensor fusion framework with cognitive bandwidth utilization. The multi-sensor data attenuation model using radio, acoustic, and visible light signal is first derived using a sum of exponentials model. A compartmental model that selectively combines the multi-sensor data is then developed. The selection of individual sensor data is based on the criterion of bandwidth utilization. The parameters of the compartmental model are computed using the modified Prony estimator, which results in high tracking accuracies. Additional advantages of the proposed method include lower computational complexity and asymptotic distribution of the estimator. Cramer-Rao bound and elliptical error probability analysis are also discussed to highlight the advantages of the compartmental model. Experimental results for real-time node tracking in indoor environment indicate a significant improvement in tracking performance when compared to state-of-the-art methods in literature.