Algorithm 1 Calculating the AMoF, fitted slope, confidence intervals, and detection threshold

1: Input: $CC{I}^{\left(D{S}_m\right)}{}_1$,………,$CC{I}^{\left(D{S}_m\right)}{}_N$, $\forall m\in n$

2: Output: AMoF, fitted slope ($\beta$), detection threshold ($\delta$)

3: At the super node

4: $\forall node\in NUT$ where the number of elements in NUT = l

5: Receive $CC{I}^{\left(S_n\right)}{}_1$,………, $CC{I}^{\left(S_n\right)}{}_N$ S.T $CC{I}^{\left(NUT\right)}isN\times n$

6: Initialize $Tem{p}^{\left(S_j\right)}$, Norm_$Tem{p}^{\left(S_j\right)}{}_i$, $AMo{F}^{\left(S_j\right)}{}_i$

7: For i = 1 to $N$ do

8: For j = 1 to n do

9: $TEM{P}_i{}^{\left(S_j\right)}$←$\left|CC{I}_{i+1}{}^{\left(S_j\right)}-CC{I}_i{}^{\left(S_j\right)}\right|+TEM{P}_{i-1}{}^{\left(S_j\right)}$

101: $Norm\_TEM{P}_i{}^{\left(S_j\right)}\leftarrow$ $TEM{P}_i{}^{\left(S_j\right)}/100$

11: End for

12: $AMo{F}_i{}^{\left(S_j\right)}$← $Norm\_TEM{P}_i{}^{\left(S_j\right)}/n+AMo{F}_i{}^{\left(S_j\right)}$

13: End for

14: Receive $AMo{F}^{\left(NUT\right)}{}_1$,..,$AMo{F}^{\left(NUT\right)}{}_{N-1}$ S.T $AMo{F}^{\left(NUT\right)}$ is $l\times \left(N-1\right)$

15: For k = 1 to $N-1$ do

16: For j = 1 to l do

17: If k ≥ 2 then

18: Find ${\beta }_k$ by solving (1)

19: Find $C_k$ by solving (2)

20: Find time varying threshold ${\delta }_k$ = $\left(\frac{max\left(C_k\right)-min\left(C_k\right)}{2}\right)+min\left(C_k\right)$

21: $@\mathit{k}=\mathbf{3}$

22: $\delta \leftarrow {\delta }_k$

23: If ${\delta }_k>\delta$

24: Node is normal

25: Else

26: Node is malicious

27: End for

28: End for