Ballistic missile tracking using the interacting mulitple model joint probabilistic data association filter
Timothy M. Dunton · 2013
Abstract : The success of interceptors used by the United States ballistic missile defense program is jeopardized by the use of hostile missile decoy deployment and evasive maneuvers. The ability to discriminate between legitimate threats and decoys is a crucial requirement for interceptor algorithms. The feasibility of the interacting multiple model joint probabilistic data association filter to effectively track a ballistic missile and detect decoys and maneuvers is the focus of this thesis. Model development and data association schemes are discussed along with optimized values for selected parameters. Performance comparisons of the resultant algorithm to a standard Kalman filter utilizing a nearest neighbor discriminator are conducted. Scenarios include combinations of missile maneuver and decoy deployment. While the Kalman filter experiences limited success, the proposed filter tracks the missile in every scenario.