Detection performance of MIMO radars in presence of K-distributed clutter using multi-dimensional saddle point approximation

Multiple Input Multiple Output (MIMO) radar is a useful tool to improve detection performance by introducing additional degree of freedom (DOF) to conventional radars at the expense of solving multi-dimensional random vectors calculation. In this paper, we develop a new multi-dimensional saddle point approximation-based approach to obtain the probability density function of multi-dimensional random vectors, which is nearly accurate especially in pdf tail regions. The proposed method is so effective to overcome the complexity of matrix calculations in determining random vector pdfs. We apply this approximated pdf to compare the detection performance of MIMO radars in the Neyman-Pearson sense against the presence of K-distributed clutter plus noise.