Joint Estimation For Interference Parameters Under Downlink Multiuser Superposition Transmission

The key idea of multiuser superposition transmission(MUST)technology is to allocate different transmission power to users through the power allocation algorithm at the transmitter side,then multiple users transmit non-orthogonally through power domain multiplexing,finally through the serial interference cancellation algorithm(SIC)to achieve the correct decoding for different user at the receiver side.In the actual transmission system,considering the signaling overhead of the system,it is impossible for the target cell base station or the neighboring cell base station to send all the transmission parameters of interfering users to the target user.Therefore,the target user needs to detect part or all of signaling of interference users,and the accuracy of blind detection will directly affect the throughput gain of non-orthogonal multiple access(NOMA)system.Firstly,this paper studies the traditional estimation algorithm for interference parameters.In the downlink MUST scenario l,a blind detection algorithm based on max-log likelihood criterion is proposed.The proposed algorithm is used to detect interference existence and interference user's modulation order under the circumstance which refers to high-order QAM composite constellation.This paper focuses on the selection of the search range enlargement factor in the proposed algorithm and obtains the detection rate curve of the blind detection algorithm corresponding to the given group of the search range enlargement factors for each candidate constellation through simulation,then further considers the impact of the signal to noise ratio(SNR)on the selection of the search range enlargement factor.Finally determines an optimal group of the search range enlargement factor for the proposed algorithm.The simulation results show that when the composite constellation for the target user and interference user is 4096QAM,the detection rate of the proposed algorithm not only reaches 100%under high SNR,but also limits the complexity of the algorithm to all below 12.5%of the total points of candidate composite constellation.Secondly,this paper studies the joint blind detection algorithm for interference existence,interference user's precoding matrix and modulation order in the downlink MUST scenario 2.This scenario is set as the target user and interference user are respectively coded and modulated on the transmitter side,then the two transmitting signals are mixed based on the fixed power allocation criterion before entering the channel,then mixed signal of the target and the interfering user is obtained at the receiver.On the basis of joint blind detection algorithm based on max-log likelihood criterion,a joint blind detection algorithm of serial interference parameters is proposed,that is to detect those interference parameters under the order of firstly detecting the precoding matrix of the interfering user(simultaneously detecting the existence of interference user)then detecting modulation order of the interfering user.In this paper,a new blind detection algorithm based on second-order transformation is proposed for blind detection of precoding matrix.The simulation results show that the blind detection accuracy of the proposed serial interference parameters joint blind detection algorithm is better than that of the parallel interference parameters joint detection algorithm over a large range of signal to noise ratio,and the proposed algorithm reaches 100%detection rate under high SNR,while the calculating complexity of the serial blind detection algorithm is only one fifth of that of the parallel interference parameters joint detection algorithm.In this paper,we optimize the joint blind detection algorithm for the interference parameters in the downlink MUST scenarios.It is of great significance to increase the throughput gain of the NOMA system by increasing the accuracy of the blind detection algorithm in practical systems.