Deep Nearest Neighbor Website Fingerprinting Attack Technology

Table 1

The parameters of the DNNF network structure.

	Parameters	Search space	Selected all of value
Model	Base model	GoogleNet, RestNet, DN4, DF	DF
	Number of filters (DF)
	Block 1 [Conv1, Conv2]	[8, …, 64]	[32, 32]
	Block 2 [Conv3, Conv4]	[32, …, 128]	[64, 64]
	Block 3 [Conv5, Conv6]	[64, …, 256]	[128, 128]
	Block 4 [Conv7, Conv8]	[128, …, 512]	[256, 256]
	Normalization studies	Batch normalization; Layer normalization	layer normalization
	Activation functions	ReLU, ELU, LeakyReLU	LeakyReLU
Model	k-NN (K)	1, 2, 3, 4, 5	3
	Similarity metrics	Euclidean, cosine	Cosine
	Loss function	Cross-entropy loss	Cross-entropy loss
	Optimizer	SGD, Adam, Adamax, RMProp	Adam