作者: Lijun Chen , Na Li
DOI: 10.1109/JSAC.2015.2482098
关键词:
摘要: Speed scaling has been widely adopted in computer and communication systems, particular, to reduce energy consumption. An important question is how speed interacts with other resource allocation mechanisms such as scheduling routing. In this paper, we study the interaction of load balancing. We characterize equilibrium resulting from balancing interaction, introduce two optimal load-balancing designs, terms traditional performance metric cost-aware (in energy-aware) metric, respectively. Especially, load-balancing–speed-scaling respect schemes processor-sharing systems. Our results show that degree inefficiency at mostly bounded by heterogeneity system, but independent number servers. These provide insights understanding guiding new designs.