Abstract:  Bearing estimation and localization are important aspects in radar signal processing.Multiple­input multiple­output(MIMO)radar exploits multiple antennas to transmit diversity waveforms and exploits multiple antennas to receive the signal reflected from the targets simultaneously，which has wide application prospects.MIMO radar has many advantages，such as waveform diversity，spatial diversity，and polarization diversity，thereby dramatically improve the performances of radar systems over the conventional phased arrays.A novel joint multiple parameter estimation and three­dimensional localization method for bistatic MIMO radar system is proposed.After the eigenvalue decomposition of the array output covariance matrix，rotated factor which contain the receive angle can be obtained by construct selection matrices.Based on the estimation of signal parameter via rotation invariance technique(ESPRIT)like algorithm，the direction of arrival(DOA)can be achieved.Owing to the steering vectors which contain the transmit elevation angle and azimuth are coupled with each other，the phase mode excitation method is used to do decoupling.In addition，we estimate the transmit elevation angle and azimuth by the uniform circular array multiple signal classification(UCA­MUSIC)and uniform circular array ESPRIT(UCA­ESPRIT)algorithm respectively.Compared with UCA­MUSIC algorithm，UCA­ESPRIT provides DOA estimates in closed­form with relatively low computational effort.Finally，based on the estimated angles，the three­dimensional(3­D)coordinates of the targets are calculated.The estimate angles can be paired automatically，and avoid additional pairing procedure.For the targets in the same range cell，the proposed algorithm can identify clearly.Simulation results show that the proposed method has the high accuracy and low complexity.