| Title | A Flow Correlation Scheme Based on Perceptual Hash and Time-Frequency Feature |
| Publication Type | Conference Paper |
| Year of Publication | 2020 |
| Authors | Wang, Zhe, Chen, Yonghong, Wang, Linfan, Xie, Jinpu |
| Conference Name | 2020 IEEE 4th Information Technology, Networking, Electronic and Automation Control Conference (ITNEC) |
| Date Published | jun |
| Keywords | flow correlation, Network security, perceptual hash, Predictive Metrics, pubcrawl, Resiliency, Scalability, Time Frequency Analysis and Security, Tracing flow |
| Abstract | Flow correlation can identify attackers who use anonymous networks or stepping stones. The current flow correlation scheme based on watermark can effectively trace the network traffic. But it is difficult to balance robustness and invisibility. This paper presents an innovative flow correlation scheme that guarantees invisibility. First, the scheme generates a two-dimensional feature matrix by segmenting the network flow. Then, features of frequency and time are extracted from the matrix and mapped into perceptual hash sequences. Finally, by comparing the hash sequence similarity to correlate the network flow, the scheme reduces the complexity of the correlation while ensuring the accuracy of the flow correlation. Experimental results show that our scheme is robust to jitter, packet insertion and loss. |
| DOI | 10.1109/ITNEC48623.2020.9085211 |
| Citation Key | wang_flow_2020 |
A Flow Correlation Scheme Based on Perceptual Hash and Time-Frequency Feature