Decoding techniques in state estimation for dynamic systems withpast histories

States of dynamic models with a higher order memory are estimated using both a stack sequential decoding algorithm and the Viterbi decoding algorithm (VDA), without higher dimensional dynamic system representation. This results in memory reduction for state estimate implementation. It is found that state estimation with a stack sequential decoding algorithm is faster and more practical than the state estimation with the Viterbi decoding algorithm, even though the estimates obtained by the Viterbi decoding algorithm are superior     

Maneuvering Target Tracking with Hypothesis Testing

A new approach is proposed for maneuvering target tracking.Target motion is described by nonlinear models in a sphericalcoordinate system. States of these models are estimated byquantization, multiple hypothesis testing, and a suboptimumdecoding algorithm of information theory. This approach does notrequire linearization of nonlinear models. Hence it is superior toclassical estimation techniques, such as the extended Kalman filter.Simulation results, some of which are presented here, haveshown the superiority of the proposed approach over target trackingwith the extended Kalman filter.     

A tracking approach for targets with measurements containingjamming or clutter signals

The concept of maneuvering target tracking which is presented by K. Demirbas, (see ibid., vol.AES-23, p.757-66, 1987) is used to track maneuvering targets whose observations contain interference representing jamming or clutter signals. The resulting tracking approach produces state estimates that closely follow the actual state values, as in target tracking in a clear environment     

Distributed sensor data fusion with binary decision trees

A distributed sensor object recognition scheme that uses object features collected by several sensors is presented. Recognition is performed by a binary decision tree generated from a training set. The scheme does not assume the availability of any probability density functions, thus it is practical for nonparametric object recognition. Simulations have been performed for Gaussian feature objects, and some of the results are presented     

Maximum a posteriori approach to object recognition withdistributed sensors

The maximum a posteriori (MAP) estimation concept is applied to the problem of object recognition with several distributed sensors. It is shown that in binary object recognition the MAP object recognition also minimizes the mean-square error. Simulation results show that the performance of the MAP object recognition is, in general, at least as good as the best performance by the sensors used