Multitarget Tracking with Doppler Ambiguity

Abstract

To my mother, who has been suppOl'ting me all the time

Constant turn model for statically fused converted measurement Kalman filters

“…Due to the nonlinearity of the sensor measurements (3), generated in polar coordinates, tracking in Cartesian coordinates usually resorts to nonlinear dynamic filtering techniques such as EKF [29] and UKF [14]. The linear CPMKF [13,32] can be used to process the range and azimuth measurements and yield satisfactory estimation performance.…”

“…As shown in (14), the Cartesian state at time step k is required to determine the matrix Γ x k . In practice, the true state is not available.…”

“…In practical tracking systems, it may be computationally prohibitive especially for multiple target tracking applications. The applicable EKF and the UKF are commonly used to incorporate the position and Doppler measurements simultaneously [14] or sequentially [9,12,29]. In [31], the Doppler measurements are also used to modify the one-point initialization technique resulting in enhanced initial track quality.…”

“…The correlation between range and Doppler measurement errors [21] is also considered while using Doppler both in state estimation and data association [20]. In [14], Doppler measurements are used to discriminate between both measurements and Doppler orders. The additional Doppler measurements are shown to be beneficial for tracking with multiple nonangular sensors [8].…”

Doppler ambiguity removal and ISAR imaging of group targets with sparse decomposition

“…In general surveillance cases, radar aims to achieve long detectable range which is the primary purpose. Moreover, low pulse repetition frequency (PRF) waveforms [14] are usually transmitted to ensure long unambiguous range measurement and enough signal energy. Another advantage of low PRF waveform lies in that it can be more tolerant of poor antenna side lobes since they can be suppressed by a sensitivity time control [15].…”