Algorithm 1: WOA for computational offloading decision-making in a smart grid

Inputs: $DA{G}_{model}$—energy and computational nodes and links, population size $P_{size}$, maximum number of iterations $t_{max}$, WOA parameters $(a, A, C$), $tasks$—workload to be relocated Outputs: $X_{sol}$—Best offloading decision solution Begin 1. $P=\left\{X_i\right\}=genInitialPopulation(DA{G}_{model}, tasks$) 2. Foreach $X_i$ determine the fitness value $F\left(X_i\right)$ 3. Select $X_{best}=min F\left(X_i\right)$ 4. Set initial values for $c, A, D, t_{max}, t=0$ 5. $\mathrm{w}\mathrm{h}\mathrm{i}\mathrm{l}\mathrm{e}\left(t<t_{\mathrm{m}\mathrm{a}\mathrm{x}}\right) \mathrm{d}\mathrm{o}$ 6.   Foreach solution $X_i^{\prime }\left(t\right)$ in $P$, do 7.    Encircling phase: 8.      Calculate distance between $X_i^{\prime }\left(t\right)$ and $X_{best}\left(t\right)$ 9.      Determine position $X_i(t+1$) using $X_i^{\prime }\left(t\right)$ and $X_{best}\left(t\right)$ in relations (6) and (7) 10.     Exploitation phase: 11.        Determine position $X_i(t+1$) using $X_i^{\prime }\left(t\right)$ and $X_{best}\left(t\right)$ in relations (8) and (9) 12.     Random exploration phase: 13.        Explore new solutions $X_i(t+1$) using relation (10) 14.     Inertia-based exploration phase: 15.        Explore new solutions $X_i(t+1$) using inertia weight in relations (12) and (13) 16.   End Foreach 17.   Foreach $X_i\left(t+1\right)$, determine the fitness value $F\left(X_i\left(t+1\right)\right)$ 18.      Select $X_{best}$= $\min F\left(X_i\left(t+1\right)\right)$ 19.   If $X_{best}$ remains unchanged for several iterations do 20.      Update exploitation / exploration parameters 21.   t = t + 1 22. end while 23. Foreach $X_i(t+1)$, apply threshold mapping to convert continuous values to binary. 24. return $X_{sol}=X_{best}$ End