Nonlinear Gaussian Filter with Multi-step Colored Noise

Color noise is a special kind of often occurring in localization systems, and it more suitable than the general Gaussian white to model time dependence due delay or high-frequency sampling. This paper derives nonlinear filtering framework for multi-step colored systems using whitening techniques Bayes rule. Meanwhile, cubature rule used solve Gaussian-weighted integral proposed framework, resulting an analytic form posterior state estimate. Compared with existing algorithms, method has obvious advantages because fully takes into account noise. Finally, effectiveness algorithm are verified classical target tracking system.