Low Complexity Track Initialization and Fusion for Multi-Modal Sensor Networks.

摘要

The objective of the project is to undertake research and development in the area of wireless sensor networks to provide an effective and affordable solution in support increasing requirements for reconnaissance and surveillance. A wireless sensor network, such as unattended ground sensors (UGS) network represents a set of inexpensive sensors of various types (modes) that provide an effective and affordable solution for battlefield surveillance. For example, acoustic sensors could obtain bearing angles of targets, and proximity sensors report the presence or absence of targets within their sensing ranges. The ultimate goal of this research is to explore fundamental performance bounds that determine how well a sensor network can resolve and localize multiple targets as a function of the operating parameters such as sensor density, the threshold settings, and the spacing between the targets. The major contribution of this project is the development of the particle-based probability density (PHD) filter for binary measurements using proximity sensors.