Full-duplex (FD) wireless and phased arrays are both promising techniques that can significantly improve data rates in future wireless networks. However, integrating FD with transmit (Tx) and receive (Rx) phased arrays is extremely challenging, due to the large number of self-interference (SI) channels. In this paper, a model is proposed to minimize the self-interference without large TxBF and RxBF gain losses through optimizing the TxBF and RxBF weights. As the model involves complex numbers and is non-convex, an iterative algorithm by solving series convex model is given to obtain an approximate solution. Meanwhile, in each step of the iterative algorithm, the sub model is translated into real number model. Then in order to balance the Tx and Rx gains, a penalty term is added into the sub model of each step of the iterative algorithm. Simulation results show that, the iterative algorithm is effective and can get good results, and the penalty term can both minimize the SI and make two gains balance.

全双工 (FD) 无线和相控阵都是有前途的技术，可以显着提高未来无线网络的数据速率。然而，由于大量的自干扰 (SI) 通道，将 FD 与发射 (Tx) 和接收 (Rx) 相控阵集成起来极具挑战性。在本文中，提出了一种模型，通过优化 TxBF 和 RxBF 权重，在没有大的 TxBF 和 RxBF 增益损失的情况下最小化自干扰。由于该模型涉及复数且是非凸的，因此给出了求解级数凸模型的迭代算法以获得近似解。同时，在迭代算法的每一步中，子模型都被转化为实数模型。然后为了平衡Tx和Rx增益，在迭代算法的每一步的子模型中都加入了一个惩罚项。