I was reading this comment and reached this paragraph:
Can we assume that GFW has entered the 2.0 era, no longer analyzing specific protocols, but rather conducting data statistics through some kind of self-learning AI to cut across the most likely part of imported traffic that is wall traffic, while domestic servers like Tencent Lightweight Cloud are spared because they do not meet the condition of overseas servers.
So it made it curious to learn more about the new techniques of TLS Fingerprinting. So I went on Arxiv and searched TLS Fingerprinting. And afterwards, I found an interesting paper from @M0hammadL, et al. which was somehow close to the descriptions of the users on this channel, who report sudden blockage of their services.
Contrary to most of the current methods, Deep Packet can identify encrypted traffic and also distinguishes between VPN and non-VPN network traffic.
After an initial pre-processing phase on data, packets are fed into Deep Packet framework that embeds stacked
autoencoder and convolution neural network in order to classify network traffic. Deep packet with CNN as its classification model achieved recall of 0.98 in application identification task and 0.94 in traffic categorization task.
To the best of our knowledge, Deep Packet outperforms all of the proposed classification methods on UNB ISCX VPN-nonVPN dataset.
I haven't read the paper thoroughly yet but I hope that it can enlightens somehow the path.
And my last words to these researchers who deliberately help the governments to crack down on the free Internet, you know exactly what you are doing. Shame on you.
I was reading this comment and reached this paragraph:
So it made it curious to learn more about the new techniques of TLS Fingerprinting. So I went on Arxiv and searched TLS Fingerprinting. And afterwards, I found an interesting paper from @M0hammadL, et al. which was somehow close to the descriptions of the users on this channel, who report sudden blockage of their services.
This paper has also a Github repo here: https://github.com/M0hammadL/DeepPacket
In the Abstract we can read:
This research has been done at Information, Network, and Learning Lab of the Sharif University of Technology in Iran.
The second author, Assistant Professor Mahdi Jafari Siavoshani had also done his postdoctoral studies at the Institute of Network Coding at The Chinese University of Hong Kong.
I haven't read the paper thoroughly yet but I hope that it can enlightens somehow the path.
And my last words to these researchers who deliberately help the governments to crack down on the free Internet, you know exactly what you are doing. Shame on you.