Algorithm 1: Flow-adaptive leased channel adjustment algorithm.

Input: $\epsilon$, $\eta$, $\varphi$, M, $L_{max}$, Input: $h,m,l:1>h>m>l>0$; Input: ${\theta }_h,{\theta }_m,{\theta }_l:1>{\theta }_h>{\theta }_m>{\theta }_l>0$. Output: $L_{}$. 1:if $\frac{\epsilon +\eta +\varphi }{M}\le {\theta }_l$ then  ▹ The proportion of PU flows is at the ultra-low level. 2:  $L_1=L_{max}$ 3:  if  $L_1+\epsilon \le M$ then 4:    $L=L_1$ 5:  else 6:    $L=M-\epsilon$ 7:if ${\theta }_l<\frac{\epsilon +\eta +\varphi }{M}\le {\theta }_m$ then  ▹ The proportion of PU flows is at the low level. 8:  $L_1=⌊h·L_{max}⌋$ 9:  if $L_1+\epsilon \le M$ then 10:    $L=L_1$ 11:  else 12:    $L=M-\epsilon$ 13:if ${\theta }_m<\frac{\epsilon +\eta +\varphi }{M}\le {\theta }_h$ then  ▹ The proportion of PU flows is at the medium level. 14:  $L_1=⌊m·L_{max}⌋$ 15:  if $L_1+\epsilon \le M$ then 16:    $L=L_1$ 17:  else 18:    $L=M-\epsilon$ 19:if${\theta }_h<\frac{\epsilon +\eta +\varphi }{M}\le 1$ then  ▹ The proportion of PU flows is at the high level. 20:  $L_1=⌊l{L}_{max}⌋$ 21:  if $L_1+\epsilon \le M$ then 22:    $L=L_1$ 23:  else 24:    $L=M-\epsilon$ 25:return L