Multi-objective Task Scheduling in cloud-fog computing using goal programming approach

Fog Computing continues to extend its usage by solving cloud computing challenges about Internet of Things (IoT). Fog nodes as a processing resource, can perform tasks generated by IoT devices. IoT as a client are concerned with the timely execution of their tasks and also lower cost services, and on the other hand, they are looking for a secured task execution. In this paper, we propose a multi-objective simulated annealing (MOSA) algorithm to allocate tasks securely on the fog and cloud nodes based on deadline constraints. The Goal Programming Approach (GPA) is applied to find a compromised solution which will satisfy multiple goals. Also, regarding the distribution of IoT tasks between fog and cloud nodes, a new goal is created called access level and scheduling based on client demand. Simulation results in four low, normal, medium, and high load scenarios showing that the proposed algorithm is on average 9.5% more efficient in terms of service delay time, 87% in terms of access level control and 49.8% in terms of deadline compared to multi-objective Particle Swarm Optimization (MOPSO), multi-objective Tabu Search (MOTS), and multi-objective Moth-Flame optimization (MOMF). Also, in terms of service cost, it has obtained acceptable results close to the average of other algorithms.