Sensitivity Analysis Of Traffic Mode Recognition In Different Frequency Mobile Communication Based On 3G/4G-LTE Emulation

With the emergence of major new changes in urbanization in China,traffic activities of urban residents have become increasingly complex.Residents' pursuit of quality of service and well-being has resulted in a combination of modes of travel modes and multi-structure travel purposes,as well as urban transport.The plan puts forward higher requirements.With the popularity of 3G/4G-LTE next-generation mobile communication networks,the accuracy and breadth of mobile phone data have been refined and expanded,which has had a new impact on the investigation and processing of existing mobile phone users' travel data.This paper proposes to build a simulation platform for traffic and communication integration.By changing the frequency of signaling events in mobile communication networks,a support vector machine learning algorithm is used to identify the average speed between positioning points for traffic travel mode identification,and to evaluate the frequency of communication positioning for traffic mode identification.Sensitivity.First of all,this paper builds a traffic simulation module based on real survey data;summarizes the communication laws of 3G/4G-LTE technology,builds a communication simulation module based on TDOA algorithm,and extracts simulated mobile communication positioning data.Provides data support for mobile communication data processing and traffic mode identification.Secondly,this paper uses the trajectory of the positioning point to estimate the average travel speed between points,and obtains the speed data of the traffic pattern recognition.Combined with the support vector machine algorithm theory and the mathematical model,the search step length,the selection of feature parameters,and the model kernel function are selected in turn.Research on the optimization of parameter values and so on,thus constructing an optimal travel information recognition model.Finally,in the context of signaling frequency of different mobile communication networks,this paper uses support vector machine algorithm to identify and evaluate the four methods of walking,bicycles,buses and cars,and through the change in the accuracy of each type of traffic mode and method.To summarize the sensitivity of traffic mode recognition.The study found that the walking and bicycle changes are relatively stable,and can maintain high recognition accuracy when the positioning frequency is reduced;the accuracy of the bus and car has been reduced quickly when the positioning frequency is high,and between the two methods It is difficult to distinguish between the high frequencies of 3G/4G-LTE communications and it is still difficult to meet the needs of traffic mode identification.