AbstractIn the pulse compression radar, it is necessary to suppress the time sidelobes which can cause a false alarm so that reliability of target detection is improved. For the pulse compression system to be adaptable for radar management which is considered to be important in recent years, it is indispensable that the radar waveforms be controlled flexibly by means of pulse compression. A method to this end can be provided by the pulse compression system using several random binary sequences with maximum entropy. However, since a coded sequence generated with a probabilistic manner is used in this method, both the time sidelobes and the mismatch loss tend to increase. Therefore, it is necessary to reduce both of them. From the forementioned viewpoint, in the case that the number of stages of the pulse compression filter is set equal to the code length, a method for deriving the filter coefficients realizing the minimum value of the peak time sidelobe level is analyzed as an optimization problem in the nonlinear programming. The pulse compression characteristics obtained by this method are described.