Research On Interference Modeling In Special Wireless Communication Band For High-speed Railway

At present high-speed railway system plays a very important role in transportation system.The wireless communication bears a lot of work in the system,that is in charge of the communication between the train and the ground console.Besides,it's also responsible for exchanging message between the devices in the train.These all play a decisive role in controlling and dispatching the train,which directly affect the safety of the high speed railway system.Therefore,the quality of the wireless communication under the high-speed railway environment is important for ensuring safety traffic and providing passengers with a superior feeling of traveling.Data mining extracts useful knowledge from unstructured and semi-structured data,which is massive,missing,noisy,random and authentic.This thesis analyzes two main factors that affect the performance of high speed railway wireless communication,uses data mining method to analyze the characteristics of high speed railway wireless channel data and the interference in the wireless communication band data,combines the wireless channel modeling of high speed railway with data mining technology.Analyzing the data of the wireless channel of high speed railway by data mining method in order to extract the useful information and make the wireless communication system model of high speed railway more complete.This thesis mainly includes the following content:At first introduces the development background and significance of data mining and high-speed railway wireless channel,as well as the relationship between them.Then introduces the basic theory of the propagation characteristics of mobile wireless channel,the basic concepts of data mining,and the relationship among data mining technology,wireless channel model and the intra band interference in wireless communication.Then simulates the multipath fading channel model to explore the method of wireless channel measurement and verify the feasibility of the method,analyses how different channel parameters affect the time-domain response of multipath fading channels.Then uses the wireless channel measurement system to measure the wireless channel data of high speed railway and the data of interference in the wireless communication band.Then uses data mining method to analyze the data,which is measured by the wireless channel measurement system.Finally analyzes the data characteristic of the interference in the wireless communication band,uses time series analysis method to analyze time domain data of receiving signalsThe contribution of this thesis mainly includes the following aspects:The wireless channel modeling method which is based on traditional theory has limitations.It is difficult to get the correlation between environmental factors and the trend prediction of channel characteristics.Besides,the model constructed by statistical sampling is difficult to respond to a complete channel model.Therefore,this thesis uses the data mining technology to overcome the limitations of traditional theoretical methods.Using CART decision tree,BP neural network,and RBF neural network to classify different types of data,which includes multi-path component data,frequency domain data,and working condition-frequency offset data of wireless channel of high speed railway,in order to construct a model that can predict the working conditions with the input data.Then,using clustering to analyze the outlier points of the frequency domain and multipath data.And marking the outlier point of multipath data as multipath component,which realizes the extraction of multipath component in time domain response of the channel.