ABSTRACT
Acknowledgments ...................................................................................................................... 316 References .................................................................................................................................... 316
The ever-increasing demand in surveillance is to produce highly accurate target identi cation and estimation in real time, even for dense target scenarios and in regions of high track contention. Past surveillance sensor systems have relied on individual sensors to solve this problem; however, current and future needs far exceed single sensor capabilities. The use of multiple sensors, through more varied information, has the potential to greatly improve state estimation and track identi cation. Fusion of information from multiple sensors is part of a much broader subject called data or information fusion, which for surveillance
applications is de ned as a multilevel, multifaceted process dealing with the detection, association, correlation, estimation, and combination of data and information from multiple sources to achieve refi ned state and identity estimation, and complete and timely assessments of situation and threat.1 (A comprehensive discussion can be found in Waltz and Llinas.1) Level 1 deals with singleand multisource information, involving tracking, correlation, alignment, and association by sampling the external environment with multiple sensors and exploiting other available sources. Numerical processes thus dominate level 1. Symbolic reasoning involving various techniques from arti cial intelligence permeates levels 2 and 3.1
Within level 1 fusion, architectures for single-and multiple-platform tracking must also be considered. These are generally delineated into centralized, distributed, and hybrid architectures,2-4 each with its advantages and disadvantages. The architecture most appropriate to the current development is that of a centralized tracking, wherein all measurements are sent to one location and processed with tracks being transmitted back to the different platforms. This architecture is optimal in that it is capable of producing the best track quality (e.g., purity and accuracy) and a consistent air picture.3,4 Although this architecture is appropriate for single-platform tracking, it may be unacceptable for multiple-platform tracking for several reasons. For example, communication loading and the single-point failure problems are important shortcomings. However, this architecture does provide a baseline against which other architectures should be compared. The case of distributed data association is discussed further in Section 13.6.
