New-Text-Document.zip_convex_convex optimization

N = 20;%Number of Channels Pt = 20;%Available Power Alpha = 1/3 + (1 - 1/3) * rand(N,1);%Alpha is between 1/3 & 1 Itteration = 1; Lambda_min = 0; Lambda_star = ( max(Alpha) + min(Alpha) ) / ( 2 * log(2) ); Lambda_max = max(Alpha); Lambda_temp = Lambda_star; P_sol = 1 / ( log(2) * Lambda_star ) - 1 ./ Alpha ; for i = 1 : N %Replacing Negative Calculated Power by 0 if P_sol(i) < 0 P_sol(i) = 0; end end P_sol_opt = sum (P_sol); while ( abs( P_sol_opt - Pt ) > 0.0001 || P_sol_opt > 20 ) %Searching for Optimum Lambda if P_sol_opt > Pt Lambda_min = Lambda_star; Lambda_star = ( Lambda_star + Lambda_max ) / 2; Lambda_temp ( Itteration + 1 ) = Lambda_star; Itteration = Itteration + 1; P_sol = 1 / ( log(2) * Lambda_star ) - 1 ./ Alpha ; for i = 1 : N if P_sol(i) < 0 P_sol(i) = 0; end end P_sol_opt = sum (P_sol); else Lambda_max = Lambda_star; Lambda_star = ( Lambda_min + Lambda_star ) / 2; Lambda_temp (Itteration+1) = Lambda_star; Itteration = Itteration + 1; P_sol = 1 / ( log(2) * Lambda_star ) - 1 ./ Alpha ; for i = 1 : N if P_sol(i) < 0 P_sol(i) = 0; end end P_sol_opt = sum (P_sol); end end