Application of hybrid meta-heuristic algorithm for assigning real-time tasks to heterogeneous processors

Heterogeneous multiprocessor system has the capability for providing low cost and high performance. Nevertheless, in order to take advantage of computing power of heterogeneous system, it is essential to use an efficient scheduling algorithm for task allocation to the available processors. This research mainly focuses on the development of task assignment algorithm for heterogeneous multiprocessor system. Assigning real-time tasks in heterogeneous multiprocessor are in general a challenging problem and NP hard. In this paper, Hybrid Max-Min Ant colony optimization algorithm (H-MMAS) is proposed to solve the real-time task assignment problem in heterogeneous multiprocessor. The objective of the proposed algorithm is to get a feasible task assignment solution, and to optimize energy consumption of every feasible task assignment solution. The performance of the proposed H-MMAS algorithm has been tested for consistent and inconsistent heterogeneous multiprocessor systems. Experimental comparisons with existing Modified BPSO algorithm demonstrate the effectiveness of the proposed H-MMAS algorithm.