Tighter bound for MULTIFIT scheduling on uniform processors

We examine one of the basic, well studied problem of scheduling theory, that of nonpreemptive assignment of independent tasks on m parallel processors with the objective of minimizing the makespan. Because this problem is NP-complete and apparently intractable in general, much effort has been directed toward devising fast algorithms which find near optimal schedules. Two well-known heuristic algorithms LPT (largest processing time first) and MULTIFIT, shortly MF, find schedules having makespans within 43, 1311, respectively, of the minimum possible makespan, when the m parallel processors are identical. If they are uniform, then the best worst-case performance ratio bounds we know are 1.583, 1.40, respectively. In this paper we tighten the bound to 1.382 for MF algorithm for the uniform-processor system. On the basis of some of our general results and other investigations, we conjecture that the bound could be tightend further to 1.366