Research On Key Technologies Of Cache Management And Security For MEC

With the rapid growth of mobile device ownership and the indispensable presence of mobile devices in people’s entertainment,work and social life,mobile users are placing increasing demands on the transmission speed and quality of service of the network when accessing resources.In order to improve the quality of service of the network,a new network architecture,Mobile Edge Computing(MEC),has been proposed.Based on the principle of temporal localization,by deploying hot resources to cache nodes near the user end,the bandwidth pressure on the core network can be greatly relieved and better quality of service can be provided.However,nodes in different geographic locations have different demands on cached content,and the location information of nodes needs to be considered in order to improve the cache hit ratio.In addition,cache nodes are vulnerable to DDos attacks,but due to their limited computing resources,deploying an attack detection model requires comprehensive consideration of the model’s resource usage and detection accuracy.In this paper,the resource management and security of edge cache nodes are studied for the above network architecture and problems.First,a geolocation-based cache management method is proposed to address the problem of how to improve the hit rate of cache nodes under MEC.With Flink extended to support realtime spatial data computing capability,the popular resources within the specified range are calculated according to the demand,and the list of resources is used as the input of the subsequent reinforcement learning model,which ensures that the resources recommended to different cache nodes are with their geographic location characteristics.Through analysis and experiments,the comparison concludes that the method proposed in this paper can effectively configure the resource list of its edge cache nodes based on geographic location information,which makes the cache hit rate significantly improved compared with the traditional approach.Second,edge cache nodes are highly vulnerable to network attacks because of their openness.In this paper,we analyze the DDos attacks for their specific attacks on cache nodes,and in addition,because of the limited computing resources of cache nodes are not suitable for the deployment of attack detection models with very large parameters on cloud nodes.To address the above issues,this paper proposes a low-power DDos attack detection method for MEC.Through the TCN network,we use hole convolution to take into account both short-term and long-term characteristics of the traffic,and reduce the complexity of the model and the parameters that need to be trained,so that the method can be better adapted to the situation of insufficient computing power of edge nodes than traditional models.