Defending against Deep-Learning-Based Flow Correlation Attacks with Adversarial Examples

Algorithm 1

Traffic consensus generate algorithm.

Input: Cycle , Time , Website Groups , Traffic File , Adversarial Samples Method
Output: Traffic consensus
(1)	While %  = = 0
(2)	For website in do
(3)	If is live then
(4)	Dump from ingress and egress Tor flows
(5)	Extract traffic characteristics from
(6)	Adversarial samples  =  ()
(7)	Write S to the traffic consensus {}
(8)	Else
(9)	Continue
(10)	End if
(11)	End while