Research On Subband Filter And Peak-toaverage Ratio Suppression Based On Filtered-OFDM System

With the development of science and technology,the fifth generation(5G)mobile communication has gradually begun to enter people's lives.In order to meet the requirements of 5G ultra-high-speed transmission data rate,high reliability and low-latency connection,it is necessary to study new wireless air interface waveforms.In the past few years,various candidate waveforms have been proposed by various parties.Filtered orthogonal frequency division multiplexing(F-OFDM)is one of them.Compared with the OFDM system adopted by 4G,the F-OFDM system has the advantages of high frequency band utilization,high parameter freedom,and low out-of-band leakage,on the premise of inheriting a series of advantages of the traditional OFDM system.The design and optimization of the filter bank at both ends of the F-OFDM system,as well as the suppression of its peak-to-average ratio,have a great impact on its application in 5G communication systems.Therefore,this paper focuses on the research of subband filter banks and peak-to-average ratio suppression technology in F-OFDM systems,including the following aspects:First,the basic principles of the OFDM system are introduced.The differences and connections between the F-OFDM system and the traditional OFDM system are compared and analyzed.The specific generation process of the baseband signal of the F-OFDM system is given.The F-OFDM system is summarized as a new generation of candidate air interface waveforms.Advantages and disadvantages.The key to the F-OFDM system lies in the design of the sub-band filter bank.This article introduces several common methods of FIR filter design based on stop band attenuation,transition bandwidth and other technical indicators.In view of the F-OFDM system's flexible configuration of parameters such as subcarrier spacing,guard interval length,number of subcarriers,and modulation method according to business requirements,this paper presents a time-frequency resource allocation scheme based on dual subbands.For the design of sub-band filters,first of all,the most common window function method is studied in detail,and the advantages and disadvantages of several types of window functions are compared in simulation.Subsequently,a subband prototype filter based on Caesar window was designed and improved,and a subband prototype filter with higher stopband attenuation was designed to suppress out-of-band leakage through the characteristics of adjustable Caesar window parameters.Through simulation analysis,it has better filtering performance than traditional window function.Secondly,in order to achieve the goal of closer to the ideal frequency response function,this paper studies the neural network and its optimization and application,proposes a design method of F-OFDM subband filter based on variable rate BP neural network,and uses BP neural network to design subband filtering Through the optimization of BP neural network to accelerate the convergence speed and minimize the out-of-band leakage.The simulation results show that the sub-band filter based on the variable-rate BP neural network has better out-of-band leakage suppression capability than the traditional window function method.A major disadvantage of the OFDM system is the high peak-to-average power ratio,and the F-OFDM system also has such deficiencies because it is also a multi-subcarrier superposition.The article first introduces the significance of the peak-to-average ratio of the system.The study compares the peak-to-average ratio performance of OFDM and F-OFDM systems.Simulation shows that the peak-to-average ratio performance of the two systems is approximately the same under the premise of having the same number of subcarriers.Subsequently,several types of peak-to-average ratio suppression methods common in traditional OFDM systems were studied and optimized according to the new F-OFDM multiband system,including:(1)Shear filtering is a simple and efficient method for peak-toaverage ratio For F-OFDM signals,an iterative shear filtering algorithm through multiplexing subband filters is studied.(2)Partial sequence transmission technology is a distortion-free peak-to-average ratio suppression algorithm.To solve the problem of excessively high calculation complexity when searching for the optimal number of times in large grouping,a partial sequence transmission based on cuckoo algorithm is proposed.Technology-CS-PTS algorithm,through intelligent algorithm to reduce the number of searches and can slightly improve the peak-to-average ratio suppression performance.(3)According to the characteristics of the F-OFDM system signal,a partial sequence transmission SG-PTS algorithm using subbands as a group is proposed,which greatly reduces the calculation complexity without changing the peak-to-average ratio suppression performance,and does not require additional grouping modules to simplify system structure.(4)Based on the new SG-PTS algorithm,it is further improved,and the MSG-PTS algorithm is proposed to improve the shortage of the SG-PTS with fewer optional rotation phases.The number of optional rotation phases is increased to improve the peak-to-average ratio suppression performance.Compared with traditional algorithms,the computational complexity is reduced,and almost the same peak-to-average ratio suppression level can be achieved.(5)The joint algorithm of MSG-PTS and shear filtering is proposed,which is compatible with the probabilistic algorithm without distortion and distortion.The method is simple and efficient,and the peak-to-average ratio suppression performance is further improved under the premise of slightly increasing the system error performance.