This research paper deals with multi-target tracking in a MIMO radar system,which presents complex data that can result in correlation problems and create technical difficulties.The objective is to resolve these issue...This research paper deals with multi-target tracking in a MIMO radar system,which presents complex data that can result in correlation problems and create technical difficulties.The objective is to resolve these issues and prevent divergence in object-tracking scenarios.However,when the cross-path phenomenon occurs,the process of assigning target measurements in MIMO radar systems becomes more complicated,and the interference phenomenon can disturb the received signal and disrupt the state estimation process.We have created a new algorithm called AMC-JPDAF that is a combination of the particle filter with the adaptive Monte Carlo(AMC)method and the joint probabilistic data association filter(JPDAF).This replaces the conventional extended KALMAN filter(EKF)used in EKF-JPDAF.We incorporated an entropy calculation and resampling sub-algorithm to overcome the limitations of EKF-JPDAF,which resulted in a more accurate estimation of two crossing targets and reduced trajectory loss in various tracking scenarios.Our experiments demonstrate that AMC-JPDAF is effective in preventing possible divergence phenomena when simulating two intersecting drones tracking scenarios.We report that the coherent measurement ambiguity is resolved at the crossover point of the trajectories corresponding to each target,giving us a low trajectory loss rate of 3.9%,which is significantly better than the 18.7%and 10.8%reported by simulations that do not affect the trajectory estimation process.We employed the MATLAB software development framework to design a system that satisfies the goals initially established by AMC-JPDAF.We then validated the system’s performance by using an experimental database collected from the MIMO-FMCW 8×16 radar system.展开更多
文摘This research paper deals with multi-target tracking in a MIMO radar system,which presents complex data that can result in correlation problems and create technical difficulties.The objective is to resolve these issues and prevent divergence in object-tracking scenarios.However,when the cross-path phenomenon occurs,the process of assigning target measurements in MIMO radar systems becomes more complicated,and the interference phenomenon can disturb the received signal and disrupt the state estimation process.We have created a new algorithm called AMC-JPDAF that is a combination of the particle filter with the adaptive Monte Carlo(AMC)method and the joint probabilistic data association filter(JPDAF).This replaces the conventional extended KALMAN filter(EKF)used in EKF-JPDAF.We incorporated an entropy calculation and resampling sub-algorithm to overcome the limitations of EKF-JPDAF,which resulted in a more accurate estimation of two crossing targets and reduced trajectory loss in various tracking scenarios.Our experiments demonstrate that AMC-JPDAF is effective in preventing possible divergence phenomena when simulating two intersecting drones tracking scenarios.We report that the coherent measurement ambiguity is resolved at the crossover point of the trajectories corresponding to each target,giving us a low trajectory loss rate of 3.9%,which is significantly better than the 18.7%and 10.8%reported by simulations that do not affect the trajectory estimation process.We employed the MATLAB software development framework to design a system that satisfies the goals initially established by AMC-JPDAF.We then validated the system’s performance by using an experimental database collected from the MIMO-FMCW 8×16 radar system.
