Multiple Differential Distinguisher of SIMECK32/64 Based on Deep Learning

Algorithm 4

Training ND_am model.

	Input:
	multiple differences (Δ0, Δ1, …, Δt − 1)
	sample number N
	Output: model
(1)	TD ← (⋅); /∗initial training set∗/
(2)	K ← Random()
(3)	P 0 = Random()
(4)	for i = 1 to t do
(5)	Pi = Pt − 1 ⊕ Δi − 1
(6)	end for
(7)	for j = 0 to N − 1 do
(8)	Cj ← encrypt (Pj, Kj)
(9)	end for
(10)	for i = 0 to N − 1 do/∗set training label∗/
(11)	if i&1 = 0 then
(12)	Ci ← Random()
(13)	Yi ← 0
(14)	else
(15)	Yi ← 1
(16)	end if
(17)	end for
(18)	TD ← (X(C0…CN − 1), Y)
(19)	model ← Training with TD
(20)	return model