Fog-Assisted Deep-Learning-Empowered Intrusion Detection System for RPL-Based Resource-Constrained Smart Industries

. 2022 Dec 2;22(23):9416. doi: 10.3390/s22239416

Algorithm 1 Proposed hybrid detection framework.

1:
Input: Dataset = $D T S$
2:
Output: $B e n i g n \to 0$ , $A t t a c k 1 \to 1$ and so on.
3:
Split the $D T S$ in to $D T S_{T r a i n i n g D a t a}$ and $D T S_{T e s t i n g D a t a}$
4:
foreach layer of $D N N G R U$ do
5:
${D T S}_{T r a i n}^{'}$ = Pre-proceessing of $D T S_{T r a i n i n g D a t a}$
6:
$D N N G R U_{T r a i n i n g}$ = Train the model using ${D T S}_{T r a i n}^{'}$
7:
$Z = B i a s + W_{1} X_{1} + W_{2} X_{2} + W_{3} X_{3} + . . . . . . . . + W_{n} X_{n}$
8:
$U_{t} = (W^{z} + X_{t} + U^{z} h_{t - 1} + B_{z})$
9:
$r_{t} = σ (W^{(r)} + X_{t} + U^{(r)} h_{t - 1})$
10:
$h_{t}^{'} = tanh (W X_{t} + r_{t} ⊙ (U) h_{t - 1})$
11:
$h_{t} = Z_{t} ⊙ h_{t - 1} + (1 - Z_{t}) ⊙ h_{t}^{'}$
12:
end for
13:
$D T S_{T e s t}^{'}$ = Pre-processing of $D T S_{T e s t i n g D a t a}$
14:
$D N N G R U_{T e s t i n g}$ = Test the model using $D T S_{T e s t}^{'}$
15:
while True do
16:
$P r e d i c t A t t a c k \to D N N G R U_{T M o d e l} (D T S_{T e s t}^{'})$
17:
if the value predicted = 0 then
18:
Return Benign
19:
else
20:
Return attack type
21:
end if
22:
end while