Research Article
Randomization-Based Dynamic Programming Offloading Algorithm for Mobile Fog Computing
| Input: number of tasks N; size of tasks ; | | Output: optimal offloading decision set ; | | (1) Initialization: initialize the DP table and set the ; | | (2) for h = 1, 2, …, H do | | (3) Generate random offloading decision sets ; | | (4) Fill in the DP table according to the rules and calculate the system utility for ; | | (5) if (the filled decisions have a common set and not the same as the cells in the previous table) then | | (6) Calculate the system utility corresponding to this decision and the original decision, respectively; | | (7) Select the decision corresponding to the lower value to replace the original decision; | | (8) Renew and ; | | (9) else | | (10) Execute the following: and ; | | (11) ifor () then | | (12) Execute the following: and ; |
|