Prior Knowledge Based Transmit Nulling Method for MIMO Radar

In this paper, a multiple-input-multiple-output (MIMO) radar transmit nulling method is proposed in the presence of fast moving interference. The MIMO radar transmit waveform design is divided into two steps: transmit waveform covariance matrix optimization and transmit waveform synthesis to approximate the optimized covariance matrix. In the first step, the prior knowledge of interference is utilized to optimize covariance matrix and the derivative constraint is introduced to widen nulls. In the second step, most existing methods, e.g. cyclic algorithm (CA) and sequential iterative algorithm (SIA), synthesize the waveform with constant modulus under leastsquares criterion. As the value at the null even can be neglected in the calculation of least-squares, these methods cannot form deep enough nulls at the interference direction. Aiming at this problem, a block coordinate descent (BCD) method is proposed to optimize the initial waveform synthesized by CA. Besides, the optimized waveform is still constant modulus. Numerical results show the efficiency of the proposed method.