Toward Robust Outlier Detector for Internet of Things Applications
Raj Mani Shukla, Shamik Sengupta · 2020
Outliers in time-series Internet of Things (IoT) data are frequent and may affect performance of the applications significantly. Therefore, their detection and mitigation are two important concerns to enable applications perform efficiently. This chapter discusses various state of the art anomaly detection techniques and provides their limitations in time-series data analysis. We emphasize the need to investigate a robust method to detect anomalies in time-series data for IoT applications. The chapter portrays a novel resilient outlier detector method that has potential to detect the anomalies injected by a nefarious adversary. We further provide the different modules of the proposed method and describe how they can be coupled together. Further, chapter delineates some of the open research issues that need to be tackled in the context of the proposed outlier detection method.