A novel approach to visualize web anomaly attacks in pervasive computing environment

摘要

These days, a pervasive computing environment is a rapidly changing trend towards increasingly always-on connected computing devices in the convergence environment. In a pervasive computing environment, there are various multimedia web services and communications for various devices in order to provide interesting and invaluable information to users. Meanwhile, providing a wide variety of the web-based multimedia services and communications may cause various security threats and abnormal behaviors. In this paper, a multimedia visualization approach for pervasive computing environment is proposed which analyzes HTTP request and response header information to detect and visualize multimedia web attacks based on the Bayesian method. We conducted a few cases' experiment for the verification of the proposed approach in a real environment. The experimental results such as web attack detection visualization, scanning and password attack visualization, and attacker's position tracking visualization verify the usability of the proposed approach.