A hybrid ant algorithm for scheduling independent jobs in heterogeneous computing environments

The efficient scheduling of independent computational jobs in a heterogeneous computing (HC) environment is an important problem in domains such as grid computing. Finding optimal schedules for such an environment is (in general) an NP-hard problem, and so heuristic approaches must be used. In this paper we describe an ant colony optimisation (ACO) algorithm that, when combined with local and tabu search, can find shorter schedules on benchmark problems than other techniques found in the literature. Introduction & Motivation The efficient scheduling of independent computational jobs in a heterogeneous computing (HC) environment such as a computational grid is clearly important if good use is to be made of such a valuable resource. However, finding optimal schedules in such a system has been shown, in general, to be NP-hard (it is a generalised reformulation of SS8 from (Garey and Johnson, 1979)). Static scheduling algorithms can be used in such a system for several different requirements (Braun et al., 2001). The first, and most common, is for planning an efficient schedule for some set of jobs that are to be run at some time in the future, and to work out if sufficient time or computational resources are available to complete the run a priori. Static scheduling may also be useful for analysis of heterogeneous computing systems, to work out the effect that losing (or gaining) a particular piece of hardware, or some subnetwork of a grid for example, will have. Static scheduling techniques can also be used to evaluate the performance of a dynamic scheduling system after it has run, to check how effectively the system is using the resources available. The ant colony optimisation (ACO) meta-heuristic was first described by Dorigo (Dorigo, 1992) as a technique to solve the travelling salesman problem, and was inspired by the ability of real ant colonies to efficiently organise the foraging behaviour of the colony using external chemical pheromone trails as a means of communication. ACO algorithms have since been widely employed on many other combinatorial optimisation problems (see (Dorigo and Stützle, 2002) for a review), including several domains related to the problem in hand, such as bin packing (Levine Copyright c © 2004, American Association for Artificial Intelligence (www.aaai.org). All rights reserved. and Ducatelle, 2003) and job shop scheduling (van der Zwaan and Marques, 1999), but ACO has not previously been applied to finding good job schedules in an HC environment.