Spatial cooperation in spectrum anomaly detection may help or hurt
When Does Spatial Cooperation Help? Contamination Regimes in Decentralized Online Spectrum Anomaly Learning
Machine Learning
Summary
Detecting unusual signals in wireless spectrum can be tricky because some unwanted transmissions slowly blend into what is considered normal background noise. The authors study how sharing information between different receivers can help find these unusual signals but can also spread errors if some receivers get wrong data. They found that whether cooperation helps depends on how often the spectrum is occupied and how similar the signals are at different receivers. They also show new ways to better catch brief or hidden bad signals by combining quick checks with longer-term monitoring.
What this means in practice
- •For wireless network operators: Detect persistent rogue transmissions more reliably by balancing cooperation between receivers to avoid spreading false detections.
- •For defense communication teams: Improve detection of hidden malicious signals that evade standard gating by combining temporal smoothing with instantaneous sample checks.
Authors
Alp Berke Ardic, Erdem Koyuncu
Abstract
Learning-based spectrum monitors adapt online to benign propagation changes, but persistent rogue transmissions can gradually become incorporated into the learned background. Spatially distributed receivers provide additional evidence for detecting such anomalies, while decentralized model mixing can also spread contamination to receivers with clean measurements. We study this tradeoff using a decentralized online learner with sample gating and model sharing. We show that, without defenses, contamination propagation is governed by the occupancy rate and is independent of cross-receiver correlation. In contrast, the benefit of spatial fusion depends strongly on correlation and the detector operating point, so cooperation can either improve or degrade detection. We also show that fragmented transmissions can evade gating based on temporally smoothed statistics. Combining smoothed and per-sample gating substantially improves robustness to such intermittent contamination.