Multiple Rao-Blackwellized particle filtering for target tracking in urban environments
Phani Chavali, Arye Nehorai · 2011
We propose a new filtering algorithm for joint tracking of multiple target states and the channel state between each pair of antennas in a radar network. The problem of tracking multiple targets in complex scenarios, such as an urban environment, poses a computational challenge as standard particle filtering (SPF) requires large number of particles to obtain an accurate estimate of the high-dimensional state vector. In this paper, we develop a hybrid filter based on the combination of multiple particle filtering (MPF) and Rao-Blackwellized particle filtering (RBPF) by exploiting the structure in the state-space model. Numerical simulations show that the proposed multiple Rao-Blackwellized particle filtering (MRBPF) performs better than the SPF and the RBPF.