Sensor Registration: Detection, Tracking, and Discrimination in a Multi-Sensor, Geographically Dispersed Environment
Sensor Management (SENMAN)
The successful performance of a Ballistic Missile Defense (BMD) System depends upon the optimal exploitation of data from a network of heterogeneous and geographically dispersed sensors. The SENMAN project will develop a real-time sensor management process to achieve optimal tracking and target discrimination performance under the constraints of limited system resources. Key functions include dynamically selecting the appropriate sensor combination from a heterogeneous sensor inventory, scheduling of sensing actions, selecting of sensor modes and generating spatially adaptive detection thresholds. We will address the data association problems of irregular, time-delayed and out-of-sequence measurements resulting from time sharing sensors between multiple targets. The sensor management process also impacts sensor registration and will need to be addressed.
The figure shows an overview of the SENMAN sensor management architecture that utilizes feedback control based upon the state estimates and error covariance matrix of the target tracks to optimally select future sensor actions.
The following sensor management functions can be performed dynamically by this approach:
Selection of the current set of sensors assigned to a threat;
Scheduling of sensing actions including temporally staggered sampling;
Selection of sensor modes including frequency and waveform; and
Tuning a spatially adaptive threshold for a Bayesian detector