Deep Learning-Driven Anomaly Detection for Green IoT Edge Networks
Bibliographic record
Abstract
The widespread use of sensor devices in IoT networks imposes a significant burden on energy consumption at the network’s edge. To address energy concerns, a prompt anomaly detection strategy is required on demand for troubleshooting resource-constrained IoT devices. It enables devices to adapt their configuration according to the dynamic signal quality and transmission settings. However, obtaining accurate energy data from IoT nodes without external devices is unfeasible. This paper proposes a framework for energy anomaly detection of IoT nodes using data transmission analysis. We use a public dataset that contains peer-to-peer IoT communication energy and link quality data. Our framework first utilizes linear regression to analyze and identify the dominant features of data communication for IoT transceivers. Later, a deep neural network modifies the gradient flow to focus on the dominant features. This modification improves the detection accuracy of anomalies by minimizing the associated reconstruction error. Finally, the energy stabilization feedback provides nodes with insight to change their transmission configuration for future communication. The experimental results show that the proposed approach outperforms other unsupervised models in anomalous energy detection. It also proves that redesigning the conventional loss function by enhancing the impact of our dominant features can dramatically improve the reliability of the anomaly detection method.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".