Research On Modulation Formats And Digital Signal Processing Algorithms In Photonics-aided Millimeter Wave Systems

In recent years,with the deployment of fifth-generation mobile communications,user data demand has exploded.Fiber-optic wireless convergence has gained widespread attention as a promising solution that combines the ultra-large bandwidth and low transmission loss of optical fiber with the mobility and flexibility of wireless networks,offering high capacity,long range and high mobility.Recently,high-speed wireless millimeter-wave transmission systems have been intensively studied to achieve higher data rates.And photon-assisted millimeter wave generation technology has been proposed to solve the problem of insufficient bandwidth for electronic devices,which also promotes seamless access between fiber optic networks and wireless links.In this thesis,we focus on the signal modulation format and the digital signal processing algorithm at the receiver side to address the factors affecting the transmission performance in photogenerated millimeter-wave systems.The main research works are as follows:1.In this thesis,the simulation of the optical-generated millimeter wave transmission system has been carried out using VPI software in conjunction with MATLAB,and the millimeter wave generation has been implemented using the optical external difference beat frequency method.The wireless channel modeling has been carried out in MATLAB where small scale fading is the Rice model.2.In this thesis,the principle of probabilistic shaping technique is analyzed,and the probabilistic shaping signal transmission combined with forward error correction code is realized by probabilistic amplitude shaping scheme,and its optimized performance and parameter selection are analyzed.Geometric shaping over AWGN channels is implemented using pairwise optimization algorithms,and a joint probabilistic and geometric shaping scheme is proposed for fiber-radio channels,using a direct design scheme and a simple iterative algorithm to further optimize the transmission performance.3.For the synchronization requirements of OFDM transmission,several synchronization algorithms are compared and optimized for the training sequence in this thesis.In this thesis,we use a pilot frequency-based channel estimation method,analyze different pilot frequency distribution schemes,use LS channel estimation at the pilot frequency,compare simple interpolation and DFT-based channel estimation algorithms at the data,and improve the traditional DFT algorithm by time-domain noise reduction and joint estimation to achieve better estimation performance.Improved phase recovery and verdict detection algorithms are proposed for the geometrically shaped constellation maps used.Using the k-means clustering algorithm for the magnitude judgement of constellation points,a better judgement performance than the maximum likelihood detection can be achieved for unequal probability transmit signals,and no exact prior information is required.Phase estimation with similar performance to the BPS algorithm is achieved with lower computational complexity by the split-loop VV algorithm with weights after amplitude judgments.