C ausal O nline L earning of Safe Regions in Cloud Radio Access Networks
Kim Hammar 2 , Tansu Alpcan 2 , Emil C. Lupu 3 · 2026
Read C ausal O nline L earning of Safe Regions in Cloud Radio Access Networks by Kim Hammar 2 , Tansu Alpcan 2 , Emil C. Lupu 3 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-05280.txt.
Description
Cloud radio access networks cut costs by scaling capacity up and down on demand, but an adaptive controller that breaks a service agreement is a liability. Kim Hammar and 2 co-authors present COL in this 2026 networking paper. The method learns the network's safe operating region: the set of resource allocations and configurations under which its specification holds.
Learning runs online in two phases. First the method watches the running network passively and infers an initial safe region using causal inference and Gaussian process regression. Then it widens that region cautiously through interventional Bayesian learning, testing the edges rather than guessing them. The authors prove the learned region is safe with a chosen probability and converges to the complete safe region under standard conditions. On a live 5G testbed, COL learned quickly at low operational cost and needed up to ten times fewer samples than leading safe-learning methods.
Catalog record
| CREATOR | Kim Hammar 2 , Tansu Alpcan 2 , Emil C. Lupu 3 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-05280