Energy saving is critical for the cloud radio access networks (C-RANs), which are composed by massive radio access units (RAUs) and energy-intensive computing units (CUs) that host numerous virtual machines (VMs). We attempt to minimize the energy consumption of C-RANs, by leveraging the RAU sleep scheduling and VM consolidation strategies. We formulate the energy saving problem in C-RANs as a joint resource provisioning (JRP) problem of the RAUs and CUs. Since the active RAU selection is coupled with the VM consolidation, the JRP problem shares some similarities with a special bin-packing problem. In this problem, the number of items and the sizes of items are correlated and are both adjustable. No existing method can be used to solve this problem directly. Therefore, we propose an efficient low-complexity algorithm along with a context-aware strategy to dynamically select active RAUs and consolidate VMs to CUs. In this way, we can significantly reduce the energy consumption of C-RANs, while do not incur too much overhead due to VM migrations. Our proposed scheme is practical for a large-size network, and its effectiveness is demonstrated by the simulation results.

中文翻译：

---

高能效云无线接入网络的动态资源供应

节能对于云无线接入网络 (C-RAN) 至关重要，云无线接入网络 (C-RAN) 由托管大量虚拟机 (VM) 的海量无线接入单元 (RAU) 和能源密集型计算单元 (CU) 组成。我们试图通过利用 RAU 睡眠调度和 VM 整合策略来最小化 C-RAN 的能耗。我们将 C-RAN 中的节能问题表述为 RAU 和 CU 的联合资源供应 (JRP) 问题。由于主动 RAU 选择与 VM 合并相结合，因此 JRP 问题与特殊的装箱问题有一些相似之处。在这个问题中，项目的数量和项目的大小是相关的并且都是可调的。没有现有的方法可以直接解决这个问题。所以，我们提出了一种高效的低复杂度算法以及上下文感知策略，以动态选择活动 RAU 并将 VM 整合到 CU。通过这种方式，我们可以显着降低 C-RAN 的能耗，同时不会因虚拟机迁移而产生过多的开销。我们提出的方案适用于大型网络，仿真结果证明了其有效性。