| Title |
Research of methods for covert channel communication identification / |
| Translation of Title |
Nedeklaruotų komunikacinių kanalų identifikavimo metodų tyrimas. |
| Authors |
Jomantas, Haroldas |
| Full Text |
|
| Pages |
48 |
| Keywords [eng] |
cybersecurity, covert communication channels, covert storage channels, covert timing channels, machine learning, convolutional neural networks. |
| Abstract [eng] |
Covert communication channels are a secret way of transmitting information using computer network communications. This work focuses on the detection of such communication channels. A covert communication channel, which has been developed and not yet described in other papers, exploits the normal functions of the ISIS protocol of a router to transmit sensitive information from one security zone to a less secured zone. The work continues with the study of identification methods, focusing on machine learning and neural network models, training them, and evaluating their performance and accuracy in identifying undeclared communication channels. A convolutional neural network was developed to identify covert communication and performed extremely well. The experiments demonstrated the superiority of this type of model over traditional machine learning models and contributed to future research on the accurate and fast identification of difficult to detect covert communication channels. |
| Dissertation Institution |
Vilniaus universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2025 |
