Efficient scheduling of independent tasks using modified heuristics

Abstract

In heterogeneous computing systems MinMin and MaxMin are widely used in assigning independent tasks to processors. For N tasks to be assigned to N processors these approaches are known to run in O (KN2) time. An algorithmic improvement that asymptotically decreases the running time complexity of MinMin to O(KN logN) without affecting its solution quality is proposed in [1]. The newly proposed MinMin algorithm is combined with MaxMin, resulting in two hybrid algorithms [1]. The first hybrid algorithm address the drawback of MaxMin in solving problem instances with highly skewed cost distributions while also improving the running time performance of MaxMin. The second hybrid algorithm improves the running time performance without degrading its solution quality. The proposed algorithms are easy to implement. For the large datasets used in the various experiments, MinMin and MaxMin, as well as recent state-of-the-art heuristics, require days, weeks, or even months to produce a solution, whereas the proposed algorithms in this paper produce solutions within only two or three minutes. The new modified algorithms namely MinMax and MinMax+ are proposed and implemented. These algorithms are compared with the existing algorithms MinMin and MaxMin on single objective cases.