Research On Efficient Signal Design And Processing Of Millimeter-Wave Multi-Antenna Communication Systems

With the development of the wireless mobile communication,information and data services are growing explosively.Future communication applications are facing urgent needs such as super bandwidth,ultra-high transmission rate,ultra-high number of equipment services and so on.At present,the low-frequency spectrum has been very crowded.Millimeter wave communication has become a promising technology in 5G network and B5 G network in the future because of its rich frequency resources.The millimeter wave band is high,the wavelength is short,and the signal suffers from serious path and penetration loss.It is necessary to use a large-scale antenna array for beamforming to provide array gain to compensate for the high attenuation.However,the millimeter wave large-scale multiple input multiple output(MIMO)system under the traditional all digital array architecture faces the problems of ultra-high hardware complexity,energy consumption and cost,which makes the millimeter wave communication still far from full deployment and application.In order to give full play to the value of millimeter wave in wireless communication,it is very necessary to develop and study the low complexity hardware architecture.Analog digital hybrid array architecture is an effective solution in the current millimeter wave large-scale MIMO system.This architecture realizes low hardware cost and complexity by reducing the number of radio frequency(RF)links,but also sacrifices the system capacity and data flow.Therefore,combined with non orthogonal multiple access(NOMA)technology,it can improve the system capacity and equipment access.However,the power amplifier(PA)in the RF link of this traditional array architecture still has the problem of low power efficiency.Compared with the traditional transmitter structure in which the signal is modulated in the baseband,the load modulation transmitter modulates the signal at the RF antenna end by changing the load impedance of the antenna.Therefore,the architecture can support multiple data streams through only one RF link,while meeting the requirements of low complexity,high power efficiency and high frequency spectral efficiency.In view of this,starting from this two low complexity hardware architectures,this dissertation makes an in-depth study on three problems: beamforming optimization design,resource allocation scheme and signal processing.The main work and contributions of this dissertation are as follows:Firstly,based on the analog-to-digital hybrid array architecture,this dissertation proposes an external strong interference suppression scheme of low complexity system based on Kronecker decomposition(KD).By using the Vandermonde characteristics of millimeter wave channel array response vector,the user combined beam is optimized to be orthogonal to the interference channel,which effectively solves the problem of near far effect of cell edge users.The proposed algorithm suppresses interference from two levels: analog domain and digital domain.It is effective for any type and intensity of interference outside the system and has universality.Secondly,this dissertation proposes several efficient waveform design and resource allocation methods of NOMA system based on analog-to-digital hybrid array architecture.In the scenario of known accurate channel state information(CSI),this dissertation proposes a user grouping scheme based on agglomerative hierarchical clustering(AGNES),which fully excavates the directional characteristics of the user CSI,realizes flexible grouping and effectively improves the system capacity;In the scenario of only knowing the second-order statistical channel state information(SCSI),this dissertation proposes a generalized agglomerative hierarchical clustering algorithm(G-AGNES)user grouping algorithm;Aiming at the problem of user fairness,this dissertation defines the signal leakage to noise radio(SLNR)criterion of NOMA systems,and proposes a robust beamforming optimization algorithm based on it;In addition,aiming at the beam overlapping problem of the NOMA system based on beam selection,this dissertation proposes two joint user grouping and beam optimization design algorithms,which effectively solves the beam overlapping problem and improves the spectral efficiency and energy efficiency of the system.Thirdly,aiming at the high complexity of the PMH signal detection process in the load modulation array architecture,a low complexity detection algorithm of the single user PMH signal is proposed in this dissertation.Based on the analysis of the signal design and channel estimation requirements under the load modulation array architecture,this dissertation proposes a semi blind channel estimation algorithm based on the expectation maximization(EM)algorithm and the least square(LS)algorithm.Based on this algorithm,a low complexity signal detection algorithm based on the spatial data structure K-dimension tree(KDT)is also proposed.The proposed algorithm achieves the same bit error rate performance as the maximum likelihood(ML)detection,but greatly reduces the implementation complexity.Finally,this dissertation studies the joint signal optimization design and low complexity signal detection in the multi-user load modulation system.This dissertation explores the influence of the signal constant envelope constraint on beamforming design under the architecture of load modulation array,and proposes an enhanced singular value decomposition(E-SVD)beamforming optimization scheme,which not only suppresses the interference between multiple users,but also effectively improves the signal gain of each user.In addition,for the joint detection of multi-user high-dimensional PMH signals,a low complexity detection algorithm based on the ball tree(BT)algorithm is proposed,which solves the problem that single-user KDT algorithm is not suitable for highdimensional signal detection.To sum up,aiming at millimeter wave large-scale array antenna system,starting from the requirements of realizing high power efficiency,high spectral efficiency and low implementation complexity,this dissertation deeply discusses the optimal beamforming design theory and efficient resource allocation mechanism under the analog-to-digital hybrid architecture,and puts forward beamforming algorithm with strong interference suppression ability and efficient resource allocation algorithm of the NOMA system;The mechanism of jointly optimizing beamforming design of interference suppression and user gain in load modulation architecture is explored,and a low complexity signal detection algorithm with optimal bit error performance is proposed.The research results provide some theoretical support for the millimeter wave large-scale MIMO communication,and also have great practical reference value for the solution of the new high-capacity mobile communication technology in the future.