Study Of Novel Modulation Techniques In Ultra-High Frequency Communication Systems

In recent years,the rapid development of wireless data communication makes how to improve the wireless communication transmission capacity to be an urgent problem,and the utilization of ultra-high frequency bands such as terahertz,visible light has become an effective way to solve this problem.No matter terahertz communication or visible light communication,however,their hardware demands are different from traditional RF communication: the light-emitting diode in visible light communication requires the transmitted signal as a real positive number,and the existing technology and devices in terahertz communications cannot fully meet the demands of terahertz communication.Thus,the appropriate modulation method for the ultra-high frequency band to achieve high-speed and high-efficiency transmission has become one of the research hotspots in the field of communication.In this paper,we explore the application of space domain index modulation technique in visible light communication,and propose various adaptive precoding algorithms to enhance the performance.Meanwhile,the terahertz spatial modulation system with hardware distortion is investigated,and the detection performance is optimized by combining with machine learning techniques.The main contents of this thesis are as follows:Chapter 1 introduces the characteristics and development prospects of UHF communications,and the specificity of their hardware makes the study of their modulation techniques meaningful.Besides,this chapter provides a detailed introduction to the development and current situation of modulation technologies for visible light and terahertz communication systems.Chapter 2 introduces the principles of spatial modulation and enhanced spatial modulation techniques,and analyzes their bit error rate(BER)performance.Also,this chapter introduces the channel model of visible light communication systems and three commonly used visible light modulation techniques,including two kinds of optical orthogonal frequency division multiplexing(O-OFDM)techniques and non-DC-based-OFDM(NDC-OFDM)techniques that sacrifice antenna index to avoid clipping.Finally,the performance of each O-OFDM system is compared and simulated in this chapter.In addition,combining enhanced spatial modulation and O-OFDM,the simulation verifies that enhanced spatial modulation has better BER performance compared to spatial modulation at the same transmission rate.Chapter 3 investigates the adaptive precoding algorithm for the optical spatial modulation system.The optimization problem is established in this chapter.First,an optical spatial modulation system with two optical emitters is considered,and the closed-form solution for pulse amplitude modulation of arbitrary modulation order is derived.It is found that the complexity is decreased by reducing the error vector space.Based on this observation,a low complexity iterative(LCI)algorithm and successive convex approximation(SCA)-assisted optimization methods are proposed for systems with more optical emitters.In addition,to further exploit the degrees of freedom in the space domain,this chapter proposes an adaptive precoding scheme for different spatial modulated symbols.Simulation results show that these algorithms can significantly improve the error performance of the systems compared to the conventional systems without precoding.The introduced symbol-based SCA algorithm outperforms the matrix-based SCA in terms of BER.Chapter 4 introduces the terahertz spatial modulation system with hardware distortion,and the channel estimation and detection techniques for this system are studied.Since hardware distortion,which is usually neglected in conventional low-frequency communications,cannot be ignored in terahertz communications,we introduce a mathematical model of hardware distortion and study the mean least squares estimation and its optimal detection adapted to the terahertz spatial modulation system.In addition,to overcome the drawback of maximum likelihood detection,this chapter proposes an extreme learning-based channel estimator and detector and compares it with other neural network-based channel estimators and detectors to verify its performance.Chapter 5 summarized the full thesis and provided prospects for future research.