LiaaS: Lawful Interception as a Service

Abstract

Machine learning techniques are the key to success for big data analytics in forthcoming 5G and cloud networks. Internet Service Providers (ISPs) and mobile networks are still relying on traditional Lawful Interception (LI) mechanisms that use error prone meta data and are vulnerable to cyber-attacks. While new identity methods are used to monitor suspected end users, the major challenge is the amount of data that needs to be monitored to find the traffic of interest related to the specific targets. On the other hand, for a conversation (audio or video) between two or multiple attendees, such as a conference call or interview, extracting, briefing and classifying important information can be prone to errors and exhaustion of resources if it is done by humans. This paper proposes an intelligent, secure, fast and reliable platform called Lawful interception as a Service (LiaaS) to detect, analyze and intercept content from different media such as voice and video call. The proposed platform also extracts the minutes of conversation and the most important information from the media (audio or video) so any desired detail can be searched from it.