Dynamic Sensor Scheduling for Target Tracking in Wireless Sensor Networks With Cost Minimization Objective

Ye Yuan,Wei Yi,Wan Choi

DOI: https://doi.org/10.1109/jiot.2022.3178265

IF: 10.6

2022-10-29

IEEE Internet of Things Journal

Abstract:Wireless sensor networks (WSNs) are demonstrated to be the increasingly essential systems for various Internet of Things (IoT)-based sensing applications. This article proposes a cost-aware dynamic sensor scheduling (CADSS) framework for WSNs with multiple tasks. At its core, a system cost function is designed to quantify the expenses of the WSNs due to task executions, and a task quality function is modeled to indicate the performance of the corresponding tasks. The proposed CADSS is further formulated as an optimization problem to minimize the system cost while maintaining the desired task qualities. In this way, a comprehensive task utility evaluation methodology for self-organized WSNs is constituted. Furthermore, by modeling the posterior Cramér–Rao lower bound (PCRLB) as the task quality function and a weighted sum of the communication and sensor scheduling cost as the system cost, the CADSS is instantiated into a multitarget tracking (MTT) application. It is shown that the formulated CADSS is a nonconvex optimization problem involving two coupled binary variables that, respectively, correspond to the scheduling of sensor and cluster head. We then propose a parallel convex relation approach to solve it effectively. Numerical results verify the effectiveness of the proposed CADSS by comparing it with state-of-the-art strategies.

computer science, information systems,telecommunications,engineering, electrical & electronic