Research On Novel Technologies Of Spatial Modulation

As a new type of multi-antenna technology,spatial modulation has been widely studied because of its low RF and low complexity.The partially-activated structure of spatial modulation can effectively reduce the search space at the receiving end,which makes the system algorithm can be easily combined with machine learning algorithms,and breaks the limitations of traditional algorithms.In addition,spatial modulation is also suitable for use in visible light communication technologies because of its sparse emission structures,which can increase the system data rate.In view of the above research issues,this thesis aims to explore new techniques based on spatial modulation systems,combines machine learning methods and spatial modulation technologies,and proposes new signal detection algorithms and power allocation algorithms from the perspective of complexity optimization and performance optimization.The machine-learning-based antenna selection algorithms is optimized.Meanwhile,the visible light communication systems based on spatial modulation are also studied.The main research content of this thesis is as follows:Chapter 1 introduced the research background of the spatial modulation system.Chapter 2 introduced the spatial modulation system model,and also introduced the related performance analysis method.Then,based on the union bound method,the theoretical performance analysis of the multi-domain joint index modulation system was performed,and the general Bit Error Ratio(BER)expression was given.Chapter 3 studied the combination of spatial modulation technology and machine learning technology.In terms of signal detection,K-means clustering detectors and supervised learning detectors were introduced.On the basis of K-means clustering detectors,an improved K-means clustering detector,which selects the initial centroids based on maximizing the minimum Euclidean distance of the observations,was proposed in order to reduce the complexity of the algorithm and avoid error floor effects.In order to further improve the performance of the system,an affine propagation cluster detector was proposed based on the concept of belief propagation,which uses the responsibility matrix and the availability matrix as metrics.Compared to the blind-based detector,we also investigated the supervised learning based detector.In our research,the rotation of the constellation symbols was used to optimize the construction of the training sequence,and the length of the training sequence is effectively reduced.Moreover,we extended the machine learning method for link adaptation.In terms of antenna selection,the antenna selection algorithm based on machine learning was introduced,based on this,the feature extraction of the samples was optimized,so that the performance of the algorithm was improved.In terms of power allocation,a power allocation algorithm based on machine learning was proposed.Compared with the traditional power allocation algorithm,the complexity of the proposed algorithm is effectively reduced.Chapter 4 introduced several system models of visible light communication systems based on spatial modulation,and compared their BER performance under different system setups.Chapter 5 summarized the full thesis,and the issues to be studied in the future were given.