Multi-target Tracking by Asynchronous Sensors Based on Extended Labeled Multi-Bernoulli Filter with Doppler Measurement

Although the algorithms based on random finite set (RFS) provides a tractable solution for multi-sensor multiple targets tracking with varying target number and observations uncertainly. But in practical, this problem is more challenging when measurements are received asynchronous. This paper proposes a solution based on the extended Labeled Multi-Bernoulli (LMB) filter for multi-target tracking of asynchronous sensors with Doppler measurements. We give the formulation of the problem and derive the explicit multi-sensor updated function based on the extension of association map about target position and velocity. The asynchronous measurements are used to update the posterior density sequentially under corresponding mapping hypothesis. Our algorithm inherits the superior of the LMB filter, the performance is better than the traditional multi-sensor LMB filter. Moreover, the utilizations of the target velocity improve the accuracy of association map given Doppler measurements. The effectiveness and superior of our method are verified by the numerical example at last.