Bounded Ant Colony Algorithm Fortask Allocation on A Network of Homogeneous Processors Using a Primary Site (BTS-ACO)

Efficient scheduling of tasks for an application is considered a crucial aspect in distributed systems to achieve a superior performance. This paper presents a task scheduling algorithm base on the Ant Colony Optimization (BTS-ACO). This algorithmdepends on an initial bound on each processor to control the procedure of task allocation. Herein, the priority of tasks is to processor with the minimal load. The algorithm investigates the effect of scheduling sorted (SLoT) and random (RLoT) list of tasks. The performance of the algorithm is demonstrated by the time taken for producing effective schedules, makespan of the schedule, and load balance of the models. The results show that BTS-ACO solution with sorted list has better performance than random list.