Modern surveillance systems often utilize multiple physically distributed sensors of different types to provide complementary and overlapping coverage on targets. In order to generate target tracks and estimates, the sensor data need to be fused. While a centralized processing approach is theoretically optimal, there are significant advantages in distributing the fusion operations over multiple processing nodes. This paper discusses architectures for distributed fusion, whereby each node processes the data from its own set of sensors and communicates with other nodes to improve on the estimates. The information graph is introduced as a way of modeling information flow in distributed fusion systems and for developing algorithms. Fusion for target tracking involves two main operations: estimation and association. Distributed estimation algorithms based on the information graph are presented for arbitrary fusion architectures and related to linear and nonlinear distributed estimation results. The distributed data association problem is discussed in terms of track-to-track association likelihoods. Distributed versions of two popular tracking approaches (joint probabilistic data association and multiple hypothesis tracking) are then presented, and examples of applications are given.

No abstract

This paper discusses two issues that need to be addressed for extending a practical single sensor multitarget tracking algorithm to the case of tracking with multiple sensors. These issues correspond to the compensation of the algorithm for asynchronous reporting rates from multiple sensors, and the interpretation and implementation of an nscan approximation technique for assigning reports from multiple sensors to targets. Procedures that account for these issues and have been demonstrated using real data are also provided. -The function of a multitarget tracking algorithm is to track all targets within the surveillance volume of interest based on reports from surveillance sensors. Tracking a target typically involves the estimation of its position and velocity. Surveillance sensors provide reports on a scan-by-scan basis. The measurements contained in each sensor report depend on the type of sensor: active sensors such as a radar typically provide range, angle, and range-rate measurements; passive sensors such as an aero-acoustic or IR sensors typically provide only angle measurements.To track multiple targets, the tracking algorithm has to first correlate the reports corresponding to each target from one scan to the next. The measurements from these reports are then used to estimate the position and velocity of the target. The key problem in multitarget tracking is the correlation of the reports for each target from one scan to the next. Factors which make this correlation process difficult are the presence of multiple maneuvering targets, the presence of false reports, missed detections for targets, random initiation and termination times for targets, and sensor measurement noise. Our objective is to show how a single sensor multitarget tracking algorithm, designed to address each of these factors, has been extended to the case of tracking with multiple sensors.To satisfy this objective, we have organized the paper as follows: Section 2 will review the single sensor multitarget tracking algorithm and identify two issues which need to be addressed for the case of tracking with multiple sensors; Section 3 will discuss the first issue which is that of compensating for the asynchronous reporting rates of multiple sensors; and Section 4 will discuss the second issue which is that of correlating the multiple sensor reports for each target. These two sections will also specify procedures to account for these issues. Section 5 provides results derived from real data which demonstrate the effectiveness of these procedures.The single sensor multitarget tracking problem has been studied extensively and several algorithms have been proposed in the past [I]. Some of these algorithms, especially those used in operational systems, are based on heuristic rules formuCaptain Martin E. Liggins, I1Rome Air Development Center Griffiss Air Force Base Rome, NY lated using intuition and experience from actual surveillance scenarios. Since these algorithms are not based on a comprehensive model of the system they fail to handle situa...

No abstract

The use of information theoretics within fusion and tracking represents an interesting addition to the problem of assessing optimal track fusion performance. This paper will explore the use of information-theoretics, namely, the use of the Kullback-Leibler as a measure ofimproving on the track assignment problem.