Comparison of particle filter and histogram filter performance for passive sonar localization

This paper compares the performance of histogram and particle filters for localizing and tracking a highly maneuvering target using two widely spaced horizontal passive sonar arrays. Both filters are numerical implementations of Bayes' filter, which use recursively estimate state vectors with nonlinear update equations and non-Gaussian prior probability density functions. The histogram filter uses a grid-based approach that is analogous to midpoint rectangular integration, while the particle filter uses a direct Monte Carlo approach. Both filters are shown to successfully track a source in an example with synthetic data given sufficient computational resources. Their performance is also compared in situations where computational power is severely restricted; then the particle filter outperforms the histogram filter in this example.