Research On Mobile Edge Computing And Video Analysis Application

Mobile Edge Computing(MEC)provides cloud computing capabilities within the access network,which can offer better service combined with applications those need ultra-low latency and large calculation.In this paper,we discuss the detection and recognition models in different scenarios and develop the video analysis application with fusion model to improve the recognition accuracy based on the MEC platform-WiCloud.Test and verify the application’s performance in the service environment with low latency and high-bandwidth provided by the WiCloud.This paper mainly completes the following:(1)Researched the mobile edge computing architecture and the traditional MEC networking scheme.Researched the cooperative work of network and edge cloud based on the SDN,the management and allocation of the computing,storage and network resources within the access network based on the NFV choreography mechanism in the form of virtual machines.(2)Researched different target tracking model,proposed the model fusion according to its applicable scenarios.Based on the mobile edge computing platform-WiCloud,developed the video analysis application and verified the accuracy and robustness of the fusion model.(3)Deployed the video analysis application on WiCloud,tested the performance and stability of video analysis services.This paper mainly discussed the combination of mobile edge computing and traditional applications.Compared with the traditional cloud computing,MEC provides a better network with low latency and high bandwidth.Developed specific applications in the corresponding scenarios,verified the improvement of the fusion model and tested the performance of the application on the platform.