Power optimization with less state transition for green software defined networking

摘要

The power consumed by the Internet industry has been growing sharply in recent years, which accounts for a considerable proportion of the overall power consumption. To reduce high power consumption in the network, numerous proposals on power-saving are proposed. However, they usually reduce power consumption by dynamically sleeping and awakening the network device according to the current traffic fluctuation. When the traffic fluctuate sharply, the network device will swing between sleep and active state frequently, which consumes a considerable power and causes a series of side effects, such as packet losses, short sustainable device service life and synchronization among devices. In this paper, based on Software Defined Networking (SDN), we propose a Power Optimization with Less State Transition (POLST) framework, which mainly achieves the following two functions: (1) Real-time network monitoring with high accuracy and low overhead; (2) Power optimization with less device state transition. Then, we formulate optimization problem of minimizing the network power consumption, while considering power consumed by device state transition. To solve it, we propose a Power Aware Routing with Less State Transition (PARLST) algorithm. Finally, simulation results show that compared with the benchmarks, POLST has excellent performance in both network monitoring and network power optimization. More specifically, in the network monitoring, POLST can achieve up to 97.2% measurement accuracy with low monitoring overhead. In the power optimization, in comparison to the benchmarks, POLST can achieve up to 72.2% energy efficiency, 56.6% decrease in Switch State Transition (SST), 53.5% decrease in Link State Transition (LST).