In this article, we propose Palisade, a distributed framework for streaming anomaly detection. Palisade is motivated by the need to apply multiple detection algorithms for distinct anomalies in the same scenario. Our solution blends low latency detection with deployment flexibility and ease-of-modification. This work includes a thorough description of the choices made in designing Palisade and the reasons for making those choices. We carefully define symptoms of anomalies that may be detected, and we use this taxonomy in characterizing our work. The article includes two case studies using a variety of anomaly detectors on streaming data to demonstrate the effectiveness of our approach in an embedded setting.

在本文中，我们提出了Palisade，一种用于流式异常检测的分布式框架。Palisade的动机是需要针对同一场景中的不同异常应用多种检测算法。我们的解决方案将低延迟检测与部署灵活性和易于修改相结合。这项工作包括对设计Palisade时所做选择的详尽描述以及做出这些选择的原因。我们仔细定义了可以检测到的异常症状，并使用这种分类法来表征我们的工作。本文包括两个案例研究，这些案例使用各种异常检测器处理流数据，以证明我们的方法在嵌入式环境中的有效性。