Date of Award
Doctor of Philosophy (PhD)
Electrical and Computer Engineering
Laurie L. Joiner
Robert G. Lindquist
Mervin C. Budge
Ballistic missiles, Ballistic missile defenses, Air defenses, Radar defense networks
This dissertation presents a new method of estimating the orientation (relative to a sensing radar) of a ballistic missile object using low range-resolution Radar Cross Section (RCS) estimates. The estimation is accomplished by combining a multi-aspect feature based Hidden Markov Model (HMM) with a low fidelity RCS model of the missile object. The RCS model of the object links the true RCS to a particular orientation or aspect angle. Utilizing this relationship with the state estimation capability of a HMM, the sequence of RCS estimates made by a radar can be decoded to estimate the orientation of the missile object. This dissertation presents the development of a new Aspect Angle Estimating (AAE) HMM and the modified Viterbi algorithm for performing AAE. The modified Viterbi algorithm utilizes a uniquely developed Controlled Fidelity Reduction (CFR) in the HMM’s state space to compensate for similar looking RCS distributions at low SNR. At specific low SNR values, the fidelity is reduced by combining states into “super-states” that have similar looking observation probability distribution functions (pdfs). As demonstrated, the modified Viterbi algorithm improves the aspect angle estimation performance under low SNR conditions by up to 50% RMS error reduction as compared to using the basic HMM Viterbi algorithm.
Moody, David C., "Aspect angle estimation of roll symmetric ballistic missile objects using low-range resolution radar returns" (2015). Dissertations. 71.