Cheng, Feng; Sapegin, Andrey; Gawron, Marian; Meinel, Christoph
Proceedings of the IEEE International Symposium on Big Data Security on Cloud (BigDataSecurity‘15)
The boundary devices, such as routers, firewalls, proxies, and domain controllers, etc., are continuously generating logs showing the behaviors of the internal and external users, the working state of the network as well as the devices themselves. To rapidly and efficiently analyze these logs makes great sense in terms of security and reliability. However, it is a challenging task due to the fact that a huge amount of data might be generated for being analyzed in very short time. In this paper, we address this challenge by applying complex analytics and modern in-memory database technology on the large amount of log data. Logs from different kinds of devices are collected, normalized, and stored in the In-Memory database. Machine learning approaches are then implemented to analyze the centralized big data to identify attacks and anomalies which are not easy to be detected from the individual log event. The proposed method is implemented on the In-Memory platform, i.e., SAP HANA Platform, and the experimental results show that it has the expected capabilities as well as the high performance.