A Dynamic Scheduling Benchmark: Design, Implementation and Performance Evaluation

Games, such as Chess, Eight Queens, and Tiles Puzzle, have traditionally been used as popular benchmarks for evaluating different problem solving strategies. Many of the game benchmarks are suitable for evaluating static scheduling techniques. In such benchmarks, the scheduling phase and the execution phase (i.e. when the schedule is executed to play the game) are disjoint. The scheduling technique can be executed to compute a complete schedule prior to the execution of any moves to play the game. Due to recent interests in on-line problem solving techniques, there is a need for benchmarks which can evaluate the performance trade-offs of dynamic scheduling techniques. Many modern video and computer games can be suitable candidates for dynamic scheduling benchmarks, since they require on-line problem solving. These benchmarks and their system testbeds should be chosen and implemented such that they can accurately reveal important performance trade-offs of dynamic scheduling techniques. In this paper, we introduce a dynamic scheduling benchmark and its system testbed. This benchmark is based on an extended version of the Tetris computer game. The rules and semantics of the game were modified to lend themselves well to evaluation of discrete problem solving and optimization techniques. The system testbed is implemented in a distributed and asynchronous fashion, on a network of workstations, to reveal performance trade-offs between scheduling time, schedule quality, and problem constraints. We provide the results of a set of experiments which evaluate the benchmark and its system, and which evaluate the performance trade-offs of a set of scheduling techniques under different benchmark constraints. 1. Introduction There exists an extensive body of literature on scheduling techniques for real-time and/or distributed systems [11][12]. These techniques have been divided into two categories, namely static and dynamic. Static algorithms determine a fixed schedule off-line prior to the start of problem solving. Such approaches require complete a priori knowledge of tasks. A static schedule may lead to poor performance if a priori knowledge of tasks is not accurate. Dynamic algorithms produce schedules on line in the hope of using more up-to-date information about the tasks and the environment. Due to their on-line nature, the complexity of the dynamic algorithms directly affects their performance[4][5]. Using low-complexity algorithms