A branch-and-bound algorithm applied to optimal radar search pattern

• Published in: Aerospace science and technology Vol. 11; no. 4; pp. 279 - 288
• Main Authors: Dodin, Pierre, Minvielle, Pierre, Le Cadre, Jean-Pierre
• Format: Journal Article
• Language: English
• Published: Elsevier SAS 01.05.2007
• Subjects: Beam scheduling; Optimization; Radar search pattern; Search theory; Beam scheduling; Search theory; Radar search pattern; Optimization
• ISSN: 1270-9638; 1626-3219
• DOI: 10.1016/j.ast.2006.12.007

In the article, the radar acquisition problem, e.g. the determination of a directional energy allocation sequence, is studied. The radar search pattern goal is the detection of a moving target whose initial location is approximately known. We have turned towards the general search theory where the observer allocates indivisible search efforts while the target presence probability spreads due to its dynamics. A few years ago, a Branch and Bound algorithm was proposed to determine the optimal sequence for a conditionally deterministic target. This operational research algorithm supposes a negative exponential detection function and a one over N detection logic, meaning that the target is declared detected if it has been detected once over a horizon of N looks. We have applied it to a narrow-beam tracking radar attempting to acquire a ballistic target. Non-trivial search patterns, such as expanding-contracting spirals, are obtained.