Reliable Target Positioning in Complicated Environments Using Multiple Radar Observations

Accurate and reliable multi-target detection and localization in complicated environments are challenging due to the existence of multiple reflectors that may distract the observations. In this paper, we try to adopt the linear frequency modulated continuous wave (LFMCW) signals, in addition to the antenna array, to perform target detection and positioning. Aiming at the ghost observations which may be generated during the measurement and data fusion phases, we try to properly combine the observations from multiple radar nodes using a probabilistic data fusion framework. The positions of targets could be obtained straightforwardly using the minimum mean squared error (MMSE) estimator. The fundamental limits of target positioning are also provided in terms of the Cramer Rao Lower Bound (CRLB). Furthermore, we try to accurately identify the target of interest by characterizing its micro-Doppler effects. Both simulations and experiments are carried out to show that the ghosts could be effectively eliminated by the proposed multi-radar fusion framework. Meanwhile, the target positioning accuracy can also be obviously improved.