Interpreting Deep Learning-Based Networking Systems
Interpreting Deep Learning-Based Networking Systems
复制标题
DOI：
10.1145/3387514.3405859
复制
发表时间：
2019-10
期刊：
影响因子：
--
通讯作者：
Zili Meng;Minhu Wang;Jia-Ju Bai;Mingwei Xu;Hongzi Mao;Hongxin Hu
中科院分区：
文献类型：
--
作者：
Zili Meng;Minhu Wang;Jia-Ju Bai;Mingwei Xu;Hongzi Mao;Hongxin Hu
虽然许多基于深度学习（DL）的网络系统已经展示了上级性能，但底层的深度神经网络（DNN）仍然是黑箱，网络运营商无法解释。可解释性的缺乏使得基于DL的网络系统在实践中难以部署。在本文中，我们提出Metis，一个框架，提供了两个一般类别的网络问题，跨越本地和全球控制的可解释性。因此，Metis引入了基于决策树和超图的两种不同解释方法，将DNN策略转换为可解释的基于规则的控制器，并基于超图分析突出关键组件。我们评估Metis在两个类别的国家的最先进的DL为基础的网络系统，并表明Metis提供人类可读的解释，同时保持几乎没有性能下降。我们进一步介绍了Metis的四个具体用例，展示了Metis如何帮助网络运营商设计，调试，部署和ad-hoc调整基于DL的网络系统。
While many deep learning (DL)-based networking systems have demonstrated superior performance, the underlying Deep Neural Networks (DNNs) remain blackboxes and stay uninterpretable for network operators. The lack of interpretability makes DL-based networking systems prohibitive to deploy in practice. In this paper, we propose Metis, a framework that provides interpretability for two general categories of networking problems spanning local and global control. Accordingly, Metis introduces two different interpretation methods based on decision tree and hypergraph, where it converts DNN policies to interpretable rule-based controllers and highlight critical components based on analysis over hypergraph. We evaluate Metis over two categories of state-of-the-art DL-based networking systems and show that Metis provides human-readable interpretations while preserving nearly no degradation in performance. We further present four concrete use cases of Metis, showcasing how Metis helps network operators to design, debug, deploy, and ad-hoc adjust DL-based networking systems.