Research On Adaptive Transmission Technology In High Speed Scattering Channels

Tropospheric scattering communication is widely used in over the horizon wireless communication because of its strong anti-interference ability and high confidentiality.However,due to the complex scattering channel environment,large propagation path loss and dual selective fading of time and frequency,the reliability and stability of the system are seriously affected.Aiming at the above problems of tropospheric scattering communication,this thesis applies adaptive coding modulation technology to scattering communication.Firstly,this thesis analyzes the basic characteristics of scattering communication,and establishes the simulation channel model according to the fading characteristics of scattering channel.Secondly,using frequency switching technology,FS-SCFDE(Frequency SwitchingSingle Carrier Frequency Domain Equalization)system is proposed,which sends the data of 16 time slots to 16 different frequency bands for signal transmission,improves the frequency diversity gain of the system,and realizes high-speed transmission under scattering channel.Then,the adaptive transmission scheme under scattering channel is designed,mainly including the related technology of the scheme,MCS(Modulation and Coding Scheme)transmission scheme with multiple rates and switching control scheme based on look-up table method.Finally,the classification algorithm of machine learning is used to replace the traditional look-up table switching control scheme,which can feed back the channel quality information more accurately and improve the throughput of the system.In the design of adaptive transmission scheme,firstly,the model block diagram of adaptive transmission system under scattering channel and the frame structure of matching frequency switching technology are given.At the same time,the key technologies in the scheme are also studied,including SNR estimation,channel estimation and effective SNR mapping algorithm.Compared with the original SNR estimation algorithm,the SNR estimation algorithm based on the pilot sequence used in this thesis can effectively estimate the SNR in real time,with better estimation performance,and its NMSE is lower than that of the original algorithm;At the same time,under the comb structure of pilot sequence,linear interpolation is used for channel estimation;Finally,the average bit mutual information per bit(MMIB)mapping algorithm is selected,and the calculation process of MMIB(Mean Mutual Information per Bit)mapping algorithm and the determination of optimal tuning factor are studied in detail.In the part of multi rate MCS transmission scheme,through the comparison of transmission performance of different modulation and coding methods,eight MCS transmission schemes are finally determined,and the traditional look-up table method is used for MCS switching control.Finally,the performance of the adaptive transmission system is analyzed through simulation,which can effectively improve the system throughput while ensuring the reliable transmission of the system.Finally,aiming at the inaccurate feedback channel quality of the traditional look-up table method,this thesis applies the machine learning algorithm to the adaptive transmission system.Firstly,the principles of KNN,SVM and AdaBoost algorithms are introduced.Secondly,the three algorithms are combined with adaptive transmission,and the changes of channel environment are learned through training samples.Through high-dimensional channel quality information,the channel state is accurately fed back,and the MCS transmission scheme is adjusted in time to further optimize the system performance.Through simulation analysis,the adaptive transmission system based on machine learning can switch the MCS transmission scheme more accurately,and effectively improve the system throughput of tropospheric scattering communication.