학술논문
Name anomaly detection for ICN
Document Type
Conference
Source
2016 IEEE International Symposium on Local and Metropolitan Area Networks (LANMAN) Local and Metropolitan Area Networks (LANMAN), 2016 IEEE International Symposium on. :1-6 Jun, 2016
Subject
Language
ISSN
1944-0375
Abstract
Information leakages are one of the main security threats in today's Internet. As ICN is expected to become the core architecture for Future Internet, it is therefore mandatory to prevent this threat. This paper proves that some ICN configuration prevents information leakages via Data packets and shows that it is an open problem to prevent interest packets from carrying encoded crucial information in their names. Assuming that names in ICN will follow the current URL format commonly used in the Internet, we get the statistics of web URL based on extensive crawling experiments of main internet organizations. Then we propose a simple filtering technique based on these statistics for firewall to detect anomalous names in ICN. The experiment shows that our filtering technique recognizes 15% of names in our dataset as malicious. As the false positive rate is still high for this filter to be used in a real world operation, this work is an important step for detecting anomalous names and preventing information-leakage in ICN.