Publication

Abstract : Recently, edge computing and data analytics have been recognized as some of the key technologies to deliver low latency, high data rate services using the 5G networks. The 5G edge infrastructure is likely to cater to a mix of cellular and non-cellular devices mainly covering the Internet of things (IoT) and miniature sensor devices. The network vulnerability and threat mitigation at the edge become utmost important to protect the network from upstream attacks originating from the sensors. In mobile network deployments, the app data from multiple user devices can be collected at the mobile edge analytics server residing on the network side. These app data feeds can be processed at the edge server to dynamically build network quality maps. The quality analysis can be utilized to precisely monitor the experience of the users. Further, the serving cells or base stations serving those users within the edge can also be identified on the maps. This can help to pinpoint the real-time traffic variations per user in each cell. The user traffic trends, derived from the app-based quality analysis, can also help to identify security anomalies that may be taking place at the edge. In this paper, a novel framework to couple the traffic analysis and security monitoring at the mobile edge has been proposed with an example manual case study. With the increasing use of artificial intelligence (AI)/machine learning (ML) in automatic traffic management and optimization in 5G, such app-based framework is likely to play a key role to harden the 5G edge fabric in the future deployments.