A computational resources scheduling algorithm in edge cloud computing: from the energy efficiency of users' perspective.

Nowadays, research in the field of task offloading not only pays attention to the computing performance of computing nodes, but also pays attention to the impact of energy consumption on computing performance. The current multi-core frequency conversion processors mainly use dynamic voltage frequency scaling (DVFS) technology to reduce processor frequency and energy consumption. The most representative is ARM's big.LITTLE which is a heterogeneous processor architecture, with large cores having strong performance and small cores being more energy efficient. In this article, we aim to implement energy-saving task offloading based on the big.LITTLE processor. We regard multi-core and multi-edge clouds as multiple computing nodes with different computing capabilities and express the task offloading problem as a computing resource scheduling problem. We modeled the calculation time and energy consumption through a non-cooperative game and proved that the non-cooperative game is an accurate potential game. The theory proves that non-cooperative games have at least one pure strategy Nash equilibrium point (NEP). We designed an edge cloud multi-user computing resource scheduling algorithm based on big.LITTLE processor (MUCRS). The simulation experiment results show that the MUCRS algorithm can obtain a suitable task offloading scheme, and the time and energy cost are lower than several similar algorithms. [ABSTRACT FROM AUTHOR]