Abstract: Interference alignment (IA) is one of the promising technologies to solve the co-channel interference in wireless communication systems. It can provide users with degree of freedom that is linear to the number of users in systems. Considering that the objective function is too complicated to obtain the optimal precoding matrix in interference alignment, this paper, according to projection heory of subspace, proves an equivalent condition that the maximum desired signal can be achieved. And a Mini-norm of inner product algorithm is proposed to select the precoding matrix. Its performance is examined in a multi-user interference system by zero-forcing (ZF) and minimum-mean-square-error (MMSE) receiver respectively. Simulation results demonstrate the sum rate of proposed scheme is higher than existing schemes.