Title :
A particle filtering technique for Jump Markov Systems
Author :
Andrieu, Christophe ; Davy, Manuel ; Doucet, Arnaud
Author_Institution :
Dept. of Math., Univ. of Bristol, Bristol, UK
Abstract :
This paper presents a particle filtering strategy in order to estimate the state of Jump Markov Systems (JMS). These processes are often met in signal communications, when the Bayesian model changes with time. Our algorithm takes advantage of the structure of the process.
Keywords :
Bayes methods; Markov processes; particle filtering (numerical methods); Bayesian model; jump Markov system estimation; particle filtering strategy; signal communication; Approximation methods; Equations; Estimation; Kalman filters; Markov processes; Mathematical model; Monte Carlo methods;
Conference_Titel :
Signal Processing Conference, 2002 11th European
Conference_Location :
Toulouse
