Aiming at the problem that conventional transmit subarray partitioning reduces the aperture and increases the complexity of feed network for hybrid phased-MIMO radar, an interleaved sparse transmit subarray partitioning method is proposed in this paper and the optimal array structure is obtained by seeking the maximum output signal-to-interference-noise ratio(SINR). Firstly, a hybrid phased-MIMO radar model with interleaved subarrays is constructed. The non-convex objective function of the maximum SINR is transformed into a convex function by using sequential convex approximations(SCA) method. The convex array optimization method is used to solve the optimal array structure of the one-dimensional and two-dimensional hybrid phased-MIMO radars with the fixed number of array elements and the number of array elements as variables. Finally, simulation results show that the proposed method not only achieves the maximum SINR and lower sidelobe level, but also has higher direction-of-arrival estimation accuracy than the conventional subarray partitioning method does.