In this fork of torproject our goal is to lable traffic inside tor. after labeling tor traffic we want to impelement a specefic neural network for learning application by application traffic inside Tor network.
To build Tor from a just-cloned git repository:
sh autogen.sh && ./configure && make && make install
In this version labeling mechanism only support socks5 tunneling.
Befor starting Tor listener you should capture packets arriving with tcpdump or wireshark.
First start tor listener with sudoer previlage, because of -a flag in netstat command used in src/or/process.h file
for finding application name now using tor.
sudo tor
One of the simple way using tor you can start any application with command below
torify <application-name>
or
torsocks <application-name>
the python script in tor_labeling directory will give .pcap file from tcpdump and /tmp/lables_cell.out and
write the packet lalels in ordering of .pcap file.
Now only we can label inbound traffic. For outbound buffers we have problem for finding the exact bytes is writing in socket after SSL encryption.