Instruction Scheduling Using MAX −MIN Ant Colony Optimization

Instruction scheduling is a fundamental step for a mapping an application to a computational device. It takes a behavioral application specification and produces a schedule for the instructions onto a collection of system processing units. The objective is to minimize the completion time of the given application while effectively utilizing the computational resources. The instruction scheduling problem isNP-hard, thus it is necessary to develop effective heuristic methods in order to provide a qualitative scheduling solution. In this paper, we present a novel instruction scheduling algorithm using MAX-MIN Ant Colony Optimization approach. The algorithm utilizes a unique hybrid approach by combining the ant colony meta-heuristic with list scheduling. The algorithm uses local and global heuristics that dynamically adjust to the input application. Compared with a number of different list scheduling heuristics, our algorithm generates better results over all the tested benchmarks. Furthermore, we show that our algorithm consistently achieves a near optimal solution.