User Behavior Prediction based Adaptive Policy Pre-fetching Scheme for Efficient Network Management

In recent years, network management is commonly regarded as an essential and promising function for managing and improving the security of network infrastructures. However, as networks get faster and network centric applications get more complex, there is still significant ongoing work addressing many challenges of the network management. Traditional passive network censoring systems lack of adaptive policy pre-fetching scheme, as a result, preventing malicious behavior (such as hacker, malware etc.) is big challenging. In this paper, we propose a novel user behavior prediction based adaptive policy pre-fetching scheme for efficient network management. A newly Distributed web User Behavior Prediction model (DUBP) is introduced first to cognize and predict user behavior. It extends a distributed DHT network to fix the bottleneck in traditional Client-Server (C/S) architecture occurred by large-scale network service requesting and massive user log analyzing and calculating. Based on user behavior sensing and prediction provided by DUBP, a further Adaptive Policy Pre-fetching and Caching scheme (APPC) is addressed for fine-grained and efficient network management. Our Universal Network (UN) will employ DUBP and APPC scheme to justify its advantages in secure network service.