Localization and Detection of Multiple Attacks in Wireless Sensor Networks Using Artificial Neural Network

Table 11

Comparison of performance of the proposed MLPANN with MK-ELM using the section of the NSL-KDD dataset with hidden layers.

Class of attacks	TP rate (%)		FP rate (%)		FN rate (%)		TN rate (%)
	MK-ELM	Proposed scheme	MK-ELM	Proposed scheme	MK-ELM	Proposed scheme	MK-ELM	Proposed scheme
DoS	98.04	99.00	0.49	0.001	1.96	1.00	99.51	99.90
Probes	95.67	97.20	0.47	0.001	4.33	2.80	99.53	99.90
R2L	76.12	96.80	0.11	0.015	23.88	3.20	99.89	99.98
U2R	50.00	90.04	0.00	0.003	50.00	9.96	100.00	99.99