Algorithm 2 Cost-effective optical fronthaul design based on the GA.

Input:B, P, N,${\kappa }_n$, $C_f$, $C_o$, $C_p$, $C_p$, $C_b$, $C_a$, $C_{O\&M}$, $C_{Sr}$, E, $E_l$, $E_o$, $E_n$, $E_{\mathrm{cool}}$, maximum number of replications, population size ($P_{pop}$), maximum number of generations ($G_{max}$), number of elites in each generation ($G_E$), mutation rate ($M_{mut}$), $D_{pn}$, $D_{bp}$, $D{1}_{max}$,$D_{max}$. Output: Optimal optical fronthaul deployment, optimal TCO. 1: Calculate the number of required splitters ${\kappa }_P$ by dividing ${\kappa }_N$ by $\zeta$ 2: while Maximum number of replications not reached do 3:     Start GA to find the optimal power splitter locations: 4:     Make initial Population size ($P_{pop}$) 5:     for all $n\in {\kappa }_N$ do 6:         for all $p\in P$ do 7:            Assign each RRH with the power splitter in terms of the minimum distance 8:            Calculate the fitness of the map and the total distance 9:            Return the sorted list of fitness 10:            Determine the best chromosome index and distance 11:            Record initial population data 12:            for $G\in G_{max}$ do 13:                Get the $G_E$ 14:                Create next generation 15:                Rank the routes 16:                Use selection to get the next-generation parents 17:                Generate matching pool 18:                Produce the offspring from selected parents 19:                Apply mutation 20:                Return the new generation population 21:                Keep the $G_E$ in next generation 22:                Apply best chromosome search 23:                Store result of each generation 24:                Return result of all generations 25:                Apply the cost parameters and calculate the optimized total cost ($TCO1$). 26:            end for 27:         end for 28:     end for 29:     Start GA for P and B mapping: 30:     Make initial Population size ($P_p$) 31:     for all ${\kappa }_P\in P$ do 32:         for all $b\in B$ do 33:            Assign each power spitter with the BBU pool in terms of the minimum distance 34:            Repeat steps (8) to (24) 35:            Follow the constraints given by Eq. (10), and (11), and calculate the cost of the required number of BBUs, AWGs and OLTs. 36:            Apply the cost parameters and calculate the optimized total cost ($TCO2$) 37:         end for 38:     end for 39: end while 40: Calculate the final optimized total cost ($TCO$ = $TCO1$ + $TCO2$)