Research On Joint Coding And Modulation Technology With High Spectral Efficiency

The rapid development of wireless communication has a profound impact on people's daily lives. With the emergence of new business and new application scenarios and continuous improvement of user requirements, it can be predicted that the number of wireless devices and data transmission rate will be explosive growth in the future. Therefore,technical requirements and challenges for very high spectrum efficiency and transmission reliability are put forward by the future wireless communication systems.Coding and modulation is the core of wireless communication systems. Modulation diversity (MD) technology can significantly improve the system diversity gain without exploiting additional resources by taking advantages of the characteristics of modulation symbols.Therefore, MD has been widely studied and applied. In order to solve the shortcomings of the current research on MD and meet the future development requirements of wireless communications, in this paper, MD technique is the core of research and we focuse on the research of the novel coded modulation technology with high spectral efficiency.Through the combination and joint optimization of MD with high spectrum efficiency techniques such as multiple-input multiple-output(MIMO) and high order modulation, the transmission reliability is significantly improved in the high spectral efficiency scenes. The major contributions of this dissertation include:1. MIMO-MD system based on ideal singular value decomposition(SVD) precodingMIMO is an effective way to improve the system spectrum efficiency. When the full channel state information (CSI) can be obtained both in the transmitter and receiver, the MIMO channel capacity can be achieved by through SVD precoding. A novel spatial multiplexing MIMO-MD system based on SVD precoding is proposed, which is based on the introduction of constellation rotation and spatial Q-component interleaving to the spatial multiplexing MIMO system with ideal SVD precoding. Based on pairwise error probability (PEP) analysis, it is proved that the proposed MIMO-MD system can improve the diversity gain without loss of multiplexing gain. In particular, full diversity can be achieved in the case of two spatial streams. To minimize the upper bound of PEP, the design criterion of optimal spatial Q-component interleaver and an optimal interleaving rule are proposed. Through average mutual information (AMI) analysis, it is verified that the proposed optimal spatial component interleaver can obtain larger AMI value and affect the optimal constellation rotation angles. Based on the AMI-maximization criterion, the constellation rotation angles are jointly optimized with the optimal spatial component interleaving algorithm. As a result, the relationship between the optimal rotation angle and the coded modulation mode (modulation mode and code rate) is established. In order to approach the capacity of coded modulation (CM) system, the MIMO-MD system with iterative demapping and decoding (MIMO-MD-ID) is proposed. Based on the analysis of extrinsic information transfer (EXIT)chart, it is found that the spatial component interleaving algorithm and constellation rotation will change the characteristics of EXIT curve. In order to maximize performance gain, we focus on the joint optimization of channel coding and modulation of MIMO-MD-ID scheme. Based on the genetic algorithm, the optimization methods of binary convolution code (BCC) generation polynomials and low density parity check code(LDPC) degree distributions that match well with the optimal optimal spatial components interleaver and rotated constellation are proposed.Depending on the above research, a complete method to jointly optimize the channel encoding, rotation modulation, spatial component interleaving and SVD precoding has been formed. It significantly improves the diversity gain and coding gain of traditional spatial multiplexing MIMO system and approaches to the channel capacity. The simulation results verify the correctness of the theoretical analysis and show that the performance of joint optimized MIMO-MD system is more close to the channel capacity. Compared with the traditional spatial multiplexing MIMO systems, MIMO-MD systems can achieve significant performance gains. Moreover, the performance gain is more significant for high spectral efficiency scenarios with high modulation order and a large number of antennas.2. MIMO-MD system based on limited feedback (LF) precodingThe LF precoding only needs to feedback the precoding matrix index (PMI) to the transmitter, which can effectively reduce the feedback overhead of the CSI. A novel LF precoded MIMO-MD system is put forward by introducing the constellation rotation and the spatial Q-component interleaver to the conventional LF precoded spatial multiplexing MIMO system with linear receiver. According to the model of the proposed system, the average symbol vector error probability of MIMO-MD system with LF precoding is analyzed firstly. In order to minimize the upper bound of the average symbol vector error probability,the optimal precoding matrix selection criterion for MIMO-MD system is proposed, which is called maximization of the minimum equivalent signal-to-noise ratio (SNR) criterion. Further more, a novel precoding matrix selection criterion with low complexity for MIMO-MD system is also proposed based on the singular values of the equivalent channel matrix. The simplified criterion avoids the calculation of equivalent SNRs for all precoding matrices and all spatial streams. When the number of spatial streams M=2, the simplified criterion is equivalent to the optimal criterion that leads to the optimal performance. For the case of M>2, the simplified criterion can only bring a small performance loss compared to the optimal criterion. In addition, based on the AMI analysis, it is verified that the proposed precoding matrix selection criterion can obtain larger AMI values affect the optimal constellation rotation angles. Based on the AMI-maximization criterion, the constellation rotation angles are jointly optimized with the optimal precoding matrix selection criterion. As a result, the relationship between the optimal rotation angle and the coded modulation mode (modulation mode and code rate) is established. In order to approach the capacity of CM system, the MIMO-MD-ID system based on LF precoding is proposed. For the MIMO-MD-ID system, the optimization of LDPC that match well with the optimal precoding matrix selection criterion and rotated constellation are investigated with multiple labeling. Thus, the codebook precoding, rotated modulation and channel coding are jointly optimized to maximize the reliability of the system.The simulation results in slow fading channels coincide with the theoretical analysis and verify that the joint optimized LF precoded MIMO-MD system can achieve significant performance gain with respect to the conventional spatial multiplexing MIMO systems. The performance gain is more obvious for high spectral efficiency scenarios with a large number of spatial streams.3. Non-binary (NB) channel coding coded MD system for digital broadcasting systemsIn order to overcome the channel fading and channel erasure, MD is adopted in the single-input single-output (SISO) system of Digital Video Broadcasting-Next Generation Handheld (DVB-NGH). Moreover,DVB-NGH is the first broadcasting protocol to employ the 2*2 spatial multiplexing. To improve the performance of open-loop spatial multiplexing MIMO in the channel conditions with high correlation,MIMO-enhanced spatial multiplexing (MIMO-eSM) is used in DVB-NGH standard. Based on the analysis, we established the relationship between eSM precoding and 4-dimensional MD technology.In order to solve the limitations of MD systems based on bit-interleaved coded modulation (BICM) in current DVB-NGH, the NB-LDPC is introduced into SISO-MD and MIMO-eSM systems to further improve the transmission reliability of DVB-NGH. The CM-SISO-MD and CM-MIMO-eSM systems are proposed to ensure that the number of finite fields is equal to the order of modulation. Based on the system models of the proposed systems, CM-AMI analysis of CM-SISO-MD system and CM-MIMO-eSM system is carried out under fading channels with erasures and MIMO fading channels with high correlation, respectively.Based on AMI-maximization criterion, the constellation rotation angles of CM-SISO-MD and CM-MIMO-eSM systems are optimized and the relationship between the optimal rotation angle and the coded modulation mode (modulation mode and code rate) is established. In addition, from the information theory perspective, the maximum performance gains can be achieved by the proposed NB-LDPC coded CM systems compared to the binary LDPC coded BICM systems in current DVB-NGH standard with different channel conditions and system configurations are investigated. The theoretical analysis and simulation verify that the proposed NB-LDPC coded CM-SISO-MD and CM-MIMO-eSM systems can obtain significant performance gain compared with the binary LDPC coded BICM systems in current standard and have more robust performance under different channel conditions and system parameters.The performance gain is more significant for the high spectral efficiency scenarios with high modulation order and ill conditioned channel with high erasures and high correlation.