Hierarchical Feature Encoding in 6G in-X Subnetworks Using Information Bottleneck
Bibliographic record
Abstract
In this paper, we investigate task-oriented communication in hierarchical in-X subnetworks, where multiple in-X devices collaboratively perform inference on a shared task by offloading their individual observations. To achieve robust and efficient collaborative inference, we propose a two-stage encoding framework tailored for this setting using information bottleneck (IB) principle. In the first stage, we focus on extracting task-relevant features from each device’s observation while mitigating the impact of channel impairments caused by wireless environment dynamics and inter-cell interference. In the second stage, a feature accumulator at the in-X access point (AP) is trained to effectively integrate features from different devices by reducing redundancy and filtering out common components across partial observations. To address the computational intractability of mutual information in our formulation for the two stages of encoding, we derive tractable variational upper bounds using a variational approximation technique. Extensive experiments on image classification tasks demonstrate that our proposed method achieves significant improvements in collaborative task inference performance and exhibits strong robustness to channel condition variations. Specifically, under challenging channel conditions, our method improves inference accuracy by up to 5%, while reducing the data load offloaded to the edge server by as much as 50%, depending on the defined resemblance index of the observations, in comparison with the baseline 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 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".