Analysis and algorithms for scheduling with minimal switching activities

Switching activities are one of the important factors in power minimization. This paper studies the scheduling problem which minimizes switching activities. We show that to find the schedule which minimizes switching activities only or minimizes latency and switching activities with or without resource constraints is NP-complete. Two heuristic scheduling algorithms to minimize switching activities are presented. The algorithms use the weighted bipartite matching to find a good schedule. The experimental results show that our algorithms can produce the near-optimal schedule which uses averagely the switching activities 35% less compared to the traditional list scheduling.