We investigated 12 years DNS query logs of our campus network and identified phenomena of malicious botnet domain generation algorithm (DGA) traffic. DGA-based botnets are difficult to detect using cyber threat intelligence (CTI) systems based on blocklists. Artificial intelligence (AI)/machine learning (ML)-based CTI systems are required. This study (1) proposed a model to detect DGA-based traffic based on statistical features with datasets comprising 55 DGA families, (2) discussed how CTI can be expanded with computable CTI paradigm, and (3) described how to improve the explainability of the model outputs by blending explainable AI (XAI) and open-source intelligence (OSINT) for trust problems, an antidote for skepticism to the shared models and preventing automation bias. We define the XAI-OSINT blending as aggregations of OSINT for AI/ML model outcome validation. Experimental results show the effectiveness of our models (96.3% accuracy). Our random forest model provides better robustness against three stateof-the-art DGA adversarial attacks (CharBot, DeepDGA, MaskDGA) compared with character-based deep learning models (Endgame, CMU, NYU, MIT). We demonstrate the sharing mechanism and confirm that the XAI-OSINT blending improves trust for CTI sharing as evidence to validate our proposed computable CTI paradigm to assist security analysts in security operations centers using an automated, explainable OSINT approach (for second opinion). Therefore, the computable CTI reduces manual intervention in critical cybersecurity decision-making.

No abstract

We performed statistical analysis on the total PTR resource record (RR) based DNS query packet traffic from a university campus network to the top domain DNS server through March 14th, 2009, when the network servers in the campus network were under inbound SSH dictionary attack. The interesting results are obtained, as follows: (1) the network servers, especially, they have a function of SSH services, generated the significant PTR RR based DNS query request packet traffic through 07:30-08:30 in March 14th, 2009, (2) we calculated sample variance for the DNS query request packet traffic, and (3) the variance can change in a sharp manner through 07:30-08:30. From these results, it is clearly concluded that we can detect the inbound SSH dictionary attack to the network server by only observing the variance of the total PTR RR based DNS query request packet traffic from the network servers in the campus network. Keywords-DNS based Detection; SSH dictionary attack; SSH brute force attack

We carried out an entropy study on the DNS query traffic from the Internet to the top domain DNS server in a university campus network through January 1st to March 31st, 2009. The obtained results are: (1) We observed a difference for the entropy changes among the total-, the A-, and the PTR resource records (RRs) based DNS query traffic from the Internet through January 17th to February 1st, 2009. (2) We found the large NS RR based DNS query traffic including only a keyword "." in the total inbound DNS query traffic. (3) We also found that the unique source IP address based PTR DNS traffic entropy slightly increased, while the unique DNS query keywords based one drastically decreased in March 9th, 2009. We found a specific IP host which was an already-hijacked classical Linux PC that carried out the SSH dictionary attack to the Internet sites in March 9th, 2009. From these results,we can detect the unusual inbound NS RR based DNS traffic and the outbound SSH dictionary attacks by only watching DNS query traffic from the Internet.

No abstract