Studies Of Adaptive Self-interference Cancellation Scheme For Full-duplex System

With the continuous development and promotion of information technology,we have entered the Internet era,People can communicate quickly and reliably with every corner of the world anytime,anywhere.With the rapid rise of mobile communication technology,wireless spectrum resources are increasingly scarce.Full-duplex(FD)technology meets the requirements of high throughput and high spectral efficiency for wireless communications.However,the key challenge in achieving full-duplex communication is self-interference from its own transmit signals.This paper mainly studies the adaptive self-interference cancellation scheme to improve the overall performance of full-duplex communication.The convergence speed of the traditional full-duplex system self-interference cancellation scheme is relatively slow,and the self-interference cancellation capability is limited.In this paper,we first study a self-interference cancellation scheme based on LMS(Least Mean Square)algorithm.It uses the LMS algorithm to the signal estimation,and uses the LMS algorithm to realize the effective self-interference cancellation scheme.The simulation results show that the self-interference cancellation ability of the system is improved obviously,which verifies the feasibility of the scheme.In order to further improve the overall performance of self-interference cancellation system,this paper proposes a self-interference cancellation scheme based on improved CMA algorithm.The CMA algorithm has low computational complexity and the cost function is only related to the amplitude of the received sequence,but not to the phase,so it is insensitive to the carrier phase.The simulation results show that the scheme can eliminate the self-interference effect.Then,combined with the advantages of CMA algorithm and LMS algorithm,a new self-interference cancellation scheme based on correlation normalized constant-mode least mean square algorithm(CMA-CNLMS algorithm)is proposed.Through simulation analysis,and with the use of CMA algorithm and LMS algorithm to eliminate the program to compare.It can be concluded that the steady-state MSE performance is also outstanding when the fast convergence characteristic of the LMS algorithm is continued by the CMA-CNLMS algorithm.