PulseAugur
LIVE 14:41:19
research · [3 sources] ·
0
research

New C-MTAD-GAT model offers unsupervised anomaly detection for mobile networks

By PulseAugur Editorial · Summary by gemini-2.5-flash-lite from 3 sources

Researchers have developed a new unsupervised anomaly detection framework called C-MTAD-GAT, designed for large-scale mobile networks. This model utilizes context-aware graph attention to monitor thousands of network elements and their high-dimensional time-series data. It has demonstrated improved performance over existing methods on a public dataset and is currently deployed by a national mobile operator, providing actionable alerts without the need for labeled incidents. AI

Summary written by gemini-2.5-flash-lite from 3 sources. How we write summaries →

IMPACT Provides a scalable, unsupervised anomaly detection solution for mobile network operators, reducing reliance on manual labeling and improving alert accuracy.

RANK_REASON This is a research paper describing a new model for anomaly detection.

Read on arXiv cs.LG →

COVERAGE [3]

  1. arXiv cs.LG TIER_1 · Sara Malacarne, Eirik Hoel-H{\o}iseth, Erlend Aune, David Zsolt Bir\'o, Massimiliano Ruocco ·

    Scalable Context-Aware Graph Attention for Unsupervised Anomaly Detection in Large-Scale Mobile Networks

    arXiv:2605.00482v1 Announce Type: new Abstract: Mobile network operators must monitor thousands of heterogeneous network elements across the radio access network and the packet core, each exposing high-dimensional KPI time series. The scale and cost of incident labelling make sup…

  2. arXiv cs.AI TIER_1 · Massimiliano Ruocco ·

    Scalable Context-Aware Graph Attention for Unsupervised Anomaly Detection in Large-Scale Mobile Networks

    Mobile network operators must monitor thousands of heterogeneous network elements across the radio access network and the packet core, each exposing high-dimensional KPI time series. The scale and cost of incident labelling make supervised approaches impractical, motivating unsup…

  3. arXiv cs.LG TIER_1 · Sara Malacarne, Eirik Hoel-H{\o}iseth, Erlend Aune, David Zsolt Biro, Massimiliano Ruocco ·

    Context-Aware Graph Attention for Unsupervised Telco Anomaly Detection

    arXiv:2604.27172v1 Announce Type: new Abstract: We propose C-MTAD-GAT, an \emph{unsupervised}, \emph{context-aware} graph-attention model for anomaly detection in multivariate time series from mobile networks. C-MTAD-GAT combines graph attention with lightweight context embedding…

RELATED ENTITIES

RELATED TOPICS