A Multiphase Dynamic Deployment Mechanism of Virtualized Honeypots Based on Intelligent Attack Path Prediction

Algorithm 1

Deployment strategy prediction algorithm.

	Input: attack graph model and the defender’s action D
	Output: and model
(1)	Initialize attack graph model G
(2)	Initialize the attacker’s strategy model and action A, and the defender’s action D
(3)	Initialize replay Memory to capacity C
(4)	Initialize DQN parameters #Q (random weights ), target (weights  = ) and replay Memory to capacity N
(5)	:
(6)
(7)
(8)	:
(9)	Get the alarm
(10)	Output: or
(11)
(12)
(13)	Store transition in
(14)	Sample random minibatch of transitions from
(15)	Set : ;
(16)	Perform a gradient descent step on with respect to the network parameters ;
(17)	Every C step reset
(18)	End while
(19)	Store attack information in #After the attack information is full, it will be sent to the BP neural network to update models
(20)
(21)	Update the defender’s action D from BP Network
(22)	End for
(23)	Output: