Constellation Division Based Space-time Modulation Technology

With the rapid increase of intelligent mobile terminals,wireless spectrum resources become increasingly scarce,and the current mobile communication system will not meet the demands of the future society.Hence,the fifth generation(5G)mobile communication system raises higher standards of the spectrum efficiency,energy efficiency,number of users and the end-to-end delay for the future wireless access technologies.Since massive multiple-input multiple-output(MIMO)has high spectrum efficiency,energy efficiency and can support a large number of users,it is widely considered as the key enabling technology of the 5G communication system.By spreading the information to different space-time resources,space-time modulation can significantly improve the spectrum efficiency and the reliability of the communication systems without using extra transmission bandwidth,and hence it will be an integral part of the future 5G standard.In addition,since the cooperative relay technology is capable of providing cooperative diversity gains,it also attracts much attention.Motivated by these facts,this dissertation focuses on the investigation of the spacetime modulation,and its application in massive MIMO and cooperative relay communication systems where constellation division is also used to eliminate the interference between different users.The main contributions of this dissertation are summarized as follows:1.Notice that the existing works on MIMO cooperative cognitive radio networks require the secondary transmitter to know the channel state information(CSI)of all the users to perform cooperative transmission with beamforming technology.However,beamforming can not efficiently suppress the interference between different users that have highly correlated CSI.Then,a novel cooperative relaying scheme called UFCPCCRN is proposed in this dissertation based on uniquely factorable constellation pair(UFCP)and space-time modulation.In the UFCP-CCRN,the primary and secondary transmitters use the two sub-constellations of the UFCP to transmit information,and the secondary transmitter employs space-time modulation scheme to amplify-andforward the information from primary user along with its own information,and then both of the primary and secondary receivers can uniquely detect their own information if the corresponding CSIs are available at both receivers.In addition,the pairwise error probabilities(PEPs)of the maximum likelihood(ML)detector are derived for both of the two receivers,and showes that the proposed scheme can achieve the full diversity in moderate and high SNR regimes.Besides,in order to minimize the PEP of primary user,a power loading algorithm is proposed under the constraint that the sum of the transmitting power of the two transmitters is limited.Simulation results demonstrate that the proposed UFCP-CCRN achieves the full diversity and outperforms the beamforming method in high SNR regions.2.On the interference elimination problem between the users that simultaneously transmit signals in the same frequency in the uplink of massive MIMO systems,this dissertation investigates the multi-user space-time modulation and the noncoherent detection while the large scale fading coefficients are known.Firstly,using the asymptotical orthogonality between different channels in massive MIMO systems,a general multi-user space-time modulation scheme is proposed based on the uniquely decomposable constellation group(UDCG).Then,an invertible two-user space-time modulation scheme is designed based on a newly defined uniquely factorable constellation(UFC).Further,an invertible three-user space-time modulation scheme is developed by the division of a cross 8-quadrature amplitude modulation(QAM).By the design of invertible space-time modulation matrices,the receiver can implement channel estimation after the signal is correctly detected.In addition,a method to develop an UDCG with Lagrange four-square theorem is also introduced.Finally,four noncoherent detectors are studied that include ML detector,minimum Riemannian distance(MRD)detector,minimum Euclidean distance(MED)detector and correlation based detector.The performance of the four detectors decrease one by one,but the required information also decreases.Therefore,we can chose different detectors depending on the channel conditions,targeting performance and computational complexity.3.Aiming to solve the interference elimination problem between the users that simultaneously transmit signals with the same frequency in the fast varying uplink of massive MIMO systems,this dissertation studies the two-user space-time modulation design and the corresponding noncoherent detection without any CSI available at the transmitters and the receiver.First,an orthogonal binary modulation division based two-user space-time modulation is proposed,and the signal matrix is invertible.Then,a fast noncoherent detector is derived based on the ML detector.In order to show the influence of large scale fading coefficients and the number of the BS antennas on the detection performance,an exact expression of the average bit error rate(BER)of the system is derived while the large scale fading coefficients are assumed to be known.In what follows,an optimization problem of power loading between two users and the signals is investigated by minimizing a tight upper bound of the average BER,and an explicit solution is obtained.Furthermore,the BS can also carry out channel estimation after the signal has been correctly detected since the space-time modulation is invertible.For the same reason,the space-time modulation can also be expanded to multiple slots conditions.4.In the fast varying channel models,if the CSI is absolutely unknown to the transmitters and the receiver,the two-user uniquely factorable space-time modulation and noncoherent detection method in the uplink massive MIMO systems is further investigated.Then,an invertible space-time modulation signal matrix is developed based on arbitrary M-ary QAM constellation division.The unique factorability of the space-time modulation is proved when the number of the BS antennas goes to infinity,and three noncoherent detectors are investigated while the number of the BS antennas is given,including the closed-form solution,generalized minimum Euclidean distance detector and the generalized minimum Riemannian distance detector.Furthermore,power loading optimization is also carried out while the distribution of the large scale fading coefficients are assumed to be known and are subject to log-normal distribution.In this case,an efficient method is obtained.Similarly,the correctly detected signal matrix can be used to perform channel estimation since the space-time modulation is invertible,and the two slots space-time modulation can also be expanded to multiple slots situations.Through the research on the multi-user space-time modulation and noncoherent detection for the uplink massive MIMO systems in this dissertation,a conclusion is obtained that the space-time modulation with constellation division and the noncoherent detection have the advantages of low delay and are ideal for the fast varying channel models where the minimum required channel coherence time is two symbol intervals.