A Scalable Approach to Solving Simulation-Based Network Security Games
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911521559281664 |
|---|---|
| author | Lanier, Michael Vorobeychik, Yevgeniy |
| author_facet | Lanier, Michael Vorobeychik, Yevgeniy |
| contents | We introduce MetaDOAR, a lightweight meta-controller that augments the Double Oracle / PSRO paradigm with a learned, partition-aware filtering layer and Q-value caching to enable scalable multi-agent reinforcement learning on very large cyber-network environments. MetaDOAR learns a compact state projection from per node structural embeddings to rapidly score and select a small subset of devices (a top-k partition) on which a conventional low-level actor performs focused beam search utilizing a critic agent. Selected candidate actions are evaluated with batched critic forwards and stored in an LRU cache keyed by a quantized state projection and local action identifiers, dramatically reducing redundant critic computation while preserving decision quality via conservative k-hop cache invalidation. Empirically, MetaDOAR attains higher player payoffs than SOTA baselines on large network topologies, without significant scaling issues in terms of memory usage or training time. This contribution provide a practical, theoretically motivated path to efficient hierarchical policy learning for large-scale networked decision problems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_16564 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | A Scalable Approach to Solving Simulation-Based Network Security Games Lanier, Michael Vorobeychik, Yevgeniy Machine Learning Cryptography and Security We introduce MetaDOAR, a lightweight meta-controller that augments the Double Oracle / PSRO paradigm with a learned, partition-aware filtering layer and Q-value caching to enable scalable multi-agent reinforcement learning on very large cyber-network environments. MetaDOAR learns a compact state projection from per node structural embeddings to rapidly score and select a small subset of devices (a top-k partition) on which a conventional low-level actor performs focused beam search utilizing a critic agent. Selected candidate actions are evaluated with batched critic forwards and stored in an LRU cache keyed by a quantized state projection and local action identifiers, dramatically reducing redundant critic computation while preserving decision quality via conservative k-hop cache invalidation. Empirically, MetaDOAR attains higher player payoffs than SOTA baselines on large network topologies, without significant scaling issues in terms of memory usage or training time. This contribution provide a practical, theoretically motivated path to efficient hierarchical policy learning for large-scale networked decision problems. |
| title | A Scalable Approach to Solving Simulation-Based Network Security Games |
| topic | Machine Learning Cryptography and Security |
| url | https://arxiv.org/abs/2602.16564 |