Abstract: An adaptive binary differential evolution (ABDE) algorithm is proposed to optimize the sparse array struc- ture of MIMO radar with constrains. The intelligent optimal algorithms always convergence in the local zone, which can be solved with the proposed approach, in certain conditions. Firstly, a canonical encoding method is proposed to cope with the constrains of arrays aperture and the limited number of elements. According to the effect of exchange elements, the arrays' structures are confirmed. Secondly, the differential evolution algorithm is improved. Referring to the harmony search algo- rithm, the random disturbing mutation is added in the proposed method to improve the diversity of populations. To conveying more information of parent to the new generation, a binary differential mutation fashion BDE/ best/1 is utilized. Then, an a- daptive adjusting strategy of coefficients to change the property of mutation is proposed, which conducts the populations to e- volve along S curve with logistic function. In the final, the effectiveness of the proposed algorithm is verified in the MIMO ra- dar arrays design.