Abstract:

A scheduling strategy to optimize energy consumption was proposed to reduce the energy consumption of the edge system in view of the fast growth of energy consumption demand in the internet of things.Based on the analysis of the power management architecture and the energy consumption composition in the typical multi-core terminal system,in view of the shortcomings of the existing system power management module,the scheduler load balancing mechanism was improved to control system energy consumption and the implementation process and implementation points of the strategy were given.The strategy was tested in the multi-core simulation system.The result shows that in low load scenarios,that is,in scenarios that CPU does not need to do data-intensive computing,the energy-saving effect of this strategy is obvious,such as data acquisition,equipment monitoring and other typical internet of things applications.

Key words: power management, scheduler, Linux