PSO Based Instruction Scheduling for Low Power

Power consumption has become an important issue in embedded system. Reducing the total number of signal transitions during each bus cycle has great effect on power consumption. In this paper, we proposed an instruction scheduling method which is based on the Particle Swarm Optimization (PSO) Algorithm to reduce the signal transitions. We formulate the low-power instruction scheduling problem to the constrained discrete PSO problem. And then we modify the PSO solution to satisfy the path constrained and find the better instruction schedule with litter signal transitions. For more effectively, we also improve the velocity updating formula to get the better result. The main contribution of our paper is that we use the improved PSO algorithm to find the better instructions sequence for low power quickly.