Network Traffic Obfuscation against Traffic Classification

Algorithm 3

Improved BIM using method 2.

	Input: classifier f; loss function J; original sample x; authentic label y; disturbance size ; fixed number of iterations n; number of iterations T.
	output: adversarial sample.
(1)	;
(2)	;
(3)	for to do
(4)	get gradient of func f with respect to ;
(5)	//update adversarial sample
(6)
(7)	if then
(8)	//modify
(9)
(10)	set all elements>0 in the vector to 0;
(11)	;
(12)	end if
(13)	end for
(14)	//according to truncate
(15)	set all elements greater than 0 in the vector to ;
	return ;