The Model Building Of OFDM System And Research On OFDM System PAPR Redection Technology

With the increasing demand for data transmission rate and transmission reliability,Orthogonal Frequency Division Multiplexing(OFDM)technology,which can greatly improve the channel capacity and transmission rate of wireless communication system and has excellent performance against multipath interference,has attracted widespread attention.This paper builds the OFDM system simulation model,analyzes the common problems of the system,and uses the corresponding algorithm for simulation test to solve these problems,and makes an in-depth study on the overall process of the system.The problem of high peak to average power ratio(PAPR)is one of the main technical obstacles of OFDM.Therefore,this paper focuses on PAPR suppression.At present,artificial intelligence technology has become increasingly mature,and is gradually used to solve various communication problems.The combination of artificial intelligence and traditional communication technology is one of the main directions of the new generation mobile communication technology.Aiming at the common problems of high complexity and large amount of computation in traditional PAPR reduction technology,this paper uses neural network.method to propose mtl-papr scheme to suppress PAPR.The innovative work of this project includes:(1)establishing mtl-papr network model based on multi task learning.If we use the traditional single task learning to deal with PAPR reduction,we need to deal with PAPR reduction and signal recovery respectively,which will ignore the relationship between the two.Therefore,it is necessary to use the idea of multi-tasking learning to build a network model,so as to achieve a network architecture model with satisfactory PAPR and BER.values and better performance;(2)SLM based algorithm The mtl-papr network model is trained by using the method of fuzzy neural network.Among the traditional algorithms for PAPR suppression,SLM algorithmhas good performance for PAPR suppression,but its computational complexity is high.As a training sample signal,it does not need to consider its acquisition speed,that is,it does not need to consider the computational complexity of the algorithm to obtain it,so it is more appropriate to use SLM algorithm with good PAPR reduction performance to obtain training data;(3)verify the effect of mtl-papr scheme.After building the model and training,the PAPR reduction performance of mtl-papr scheme is verified,and compared with the performance of traditional PAPR reduction technology.Through the comparative analysis of the simulation results,it is proved that the MTL-PAPR scheme can achieve better PAPR reduction effect with lower computational complexity under the condition of lower BER performance loss.Compared with SLM algorithm which has similar inhibition effect on PAPR,the calculation time of MTL-PAPR scheme is only about 1/15 of its.