how to implement this algorithm

hi every one , i m trying to imliment this algorithm to solve a quadratic function:

Step 1 (Initial solution)

Find an initial solution using a basic local search algorithm 

Step 2 (Tabu Search)

Repeat Steps 2.1 to 2.2 for Number Iterations iterations

2.1 (Explore the neighborhood)

2.1.1 Determine the best move while taking into
consideration the tabu moves and the aspiration
criteria. For each move x → x, we find the
solution by solving GPU(x). The cost of the
solution is the total cost of the network.

2.1.2 Determine the number of iterations (according
to a uniform distribution) for which the chosen
site is tabu.

2.2 (TS best solution update)
If the cost of the current solution is less than the cost of
the best solution found so far, update this best solution.

Step 3 (Multi-start)

3.1 (Update the solution)

If the cost of the current solution is less than the cost of
the best solution found so far, update this best solution.

3.2 (Stop condition)

If the number of start is smaller than the maximum
allowed number of starts go to step 2. Otherwise, return
the best cost found so far.

i try this:

    % Initialiser le nombre d'itérations :
 it_max=100;
 it=1;
 tic
 % génération du solution initiale
  S1=randi(SGM,1,eNodeB);
  n=length(S1);
  % Initialiser la liste de Taboue
  tL=zeros(1,n);
  % initialisé la meilleur solution 
  SB=S1;
  % calculer le cout de la solution initiale
  F=soleval(SB,SGM,eNodeB,c,h);
  % initialisé le meilleur cout
  FB=F;
  %recherche de taboue
  while(it<=it_max)
   % Explorer le voisinage
    best_cost=inf;
    for j=1:n
    if(tL(j)==0)
  SV=perm1(SB,j);
    FV=soleval1(SV,SGM,eNodeB,c,h);

    % si le cout de la solution courante <cout de meilleur solution
    % choisir le meilleur mouvement selon le cout

    if FV< best_cost
      best_cost =FV;
      Bmv=j;
    end
    end
    end 
    % mettre à jour la meilleur solution par la meilleur solution voisine  
    SB=perm1(SB,Bmv);
    % mettre à jours la liste de Taboue
    tL=tL-(tL>0);
    % determiner le nombre des itérations pour lesquelle le mouvement
    % choisit est tabou(valeur choisit aléatoirement en 5 et 9)

    tL(Bmv)=7;

    % si le cout de la solution courante <cout de meilleur solution
    % mettre à jour la meilleur solution 

    if(best_cost<=FB)
      FB=best_cost;
      SB=SV;
    end
    it=it+1;
    it
end

but, it dosent working , tha,k you for helping me.

[Walter Roberson]

What happens when you try it? Do you get an error message?

[ISIMG]

no i dont get an error message but nothing is changing,the function which i wnat to optimize is the same all the time

[ISIMG]

???

[Boudjelloul Imane]

what does mean SGM!!