Research On University LTE Network Traffic Forecast And Capacity Expansion

The launch of operators' unlimited data packages has caused explosive high data traffic growth.With the widespread promotion of unlimited traffic packages in colleges and universities,user demand has shifted from low-traffic communication services to high-traffic video services,and rapid traffic growth has consumed a lot of network resources and made campus user perceptions significantly lower.How to improve user perception to release network resources and provide higher service usage rates is the focus and difficulty of current university LTE network optimization.None of the existing traffic prediction models can effectively learn long-sequence and long-span traffic information,making the potential traffic change trend ignored,and the prediction accuracy is low,resulting in insufficient network scheduling resources.The concentration of university users,the regular schedule of work and rest,and the obvious tidal effect,the problem of limited capacity in the community has not been properly solved since the introduction of the unlimited traffic package.In view of the existing issues,the main research works in this article are as follows:(1)In terms of LTE network traffic prediction in colleges and universities,this paper proposes a prediction algorithm based on decomposition and optimization integrated learning LTE network traffic.Based on user behavior,adaptive meshing is performed on hotspot areas of the real-time network to accurately locate traffic changes in different areas,and the flow sequence is decomposed using MEMD(multivarite empirical mode decomposition)to decompose the flow sequence and strengthen the self-information of flow information.Adaptability,more detailed capture of changes in flow characteristics.The decomposed flow sequence is used to learn and predict the future flow change trend through an artificial intelligence learning algorithm to obtain the optimal solution for flow prediction.Experimental results show that using MEMD to decompose the traffic method is very effective,reducing the difficulty of traffic prediction,and can accurately predict the network traffic demand with a large span of time,providing effective guidance for the subsequent network construction of colleges and universities.(2)In terms of university LTE network expansion,this article addresses the changes in the campus user model before and after the unlimited traffic package.Based on user perception,all popular App applications are divided into web browsing,instant messaging,social networking,video,and file download.5 This type is compared with the traffic before and after the launch of unlimited traffic packages,and the frequency of use is analyzed.A least-squares method is used to establish a business traffic forecasting model for busy colleges and universities.Combined with the business guarantee rate model and the single-cell service carrying capacity,this paper proposes Method for users to perceive modest expansion of college LTE networks.The experimental results show that this method can well solve the problem of super busy network encountered by colleges and universities,and the user perception has been significantly improved,which has a certain guiding role for the future expansion and planning of colleges and universities in different scenarios.(3)In terms of data extraction of LTE network in colleges and universities,in order to solve the work difficulty of the existing network monitoring personnel and avoid the errors that may occur in the field detection network indicators,this paper builds a remote campus LTE network real-time monitoring system,remote monitoring Happening.The experiment proves that the system can monitor the network situation of different scenes in colleges and universities,and the operation is simple,the measurement result is accurate,and the practical value is high.