Open RAN Slicing for MVNOs With Deep Reinforcement Learning
Bibliographic record
Abstract
As 5G networks continue to be deployed and 6G networks begin to be envisioned, mobile network operators (MNOs) are embarking on a revolutionary transformation of the way they manage their networks. Various technology bricks are currently considered paramount in this transformation, including radio access network (RAN) slicing. The concept of an open radio access network (Open RAN) promises to provide more flexibility to support RAN slicing. However, RAN slicing in an O-RAN architecture raises a major challenge in achieving efficient resource sharing among slices, due to the diverse and permanent changes in RAN slices’ QoS requirements. To overcome this challenge in a RAN environment involving an MNO and multiple mobile virtual network operators (MVNOs), we propose a two-level RAN slicing mechanism. The first level is executed on a long time-scale to allocate radio resources from the MNO to MVNOs while the second level is executed on a shorter time-scale to allocate MVNO resources to users. This mechanism improves the performance of the RAN slicing operation by enabling users to obtain the required resources as quickly as possible and with a high level of granularity. We formulate the two-level problem as two mathematical optimization problems and we study their NP hardness. To efficiently solve the two-level problem, we first propose a game-theoretic solution to solve the first-level resource allocation problem using a matching algorithm. Next, we propose a deep reinforcement learning (DRL) algorithm that uses the double deep$Q$-network procedure to solve the second-level resource allocation problem. The two proposed algorithms are coupled such that the DRL algorithm uses the solution obtained using the game-theoretic matching algorithm. We show through extensive simulations that the proposed two-level solution outperforms the current state-of-the-art solutions and achieves efficient performance.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 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".