Anomaly Detection in Satellite Communications Networks using Support Vector Machines

Anomaly Detection in Satellite Communications Systems using LSTM Networks

“…Since it is not possible to detect novel or 'zero day' anomalies by their signatures, we [3] and others [6]- [8] have taken a different approach. Rather than using labelled anomalies, we train a model only with normal, unlabelled data, and flag any deviations from this model of normality.…”

“…

Most satellite communications monitoring tools use simple thresholding of univariate measurements to alert the operator to unusual events [1] [2]. This approach suffers from frequent false alarms, and is moreover unable to detect sequence or multivariate anomalies [3]. Here we consider the problem of detecting outliers in high-dimensional time-series data, such as transponder frequency spectra.

…”

Unsupervised anomaly detection of polluted telemetry data based on a reconstruction model with adaptive weighting

“…Therefore, attentions have been transferred to data-driven-based unsupervised methods [2], which can be further classified into clustering, classification, and reconstruction-based methods. The representative methods of the clustering-based methods and the classification-based methods are self-organizing map (SOM) [3] and one-class support vector machine (OCSVM) [4]. However, these methods fail to learn the temporal features of telemetry data and are insufficient when prone to multi-dimensional telemetry data.…”