English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Advancements in reinforcement learning (RL) have inspired new directions inintelligent automation of network defense. However, many of these advancementshave either outpaced their application to network security or have notconsidered the challenges associated with implementing them in the real-world.To understand these problems, this work evaluates several RL approachesimplemented in the second edition of the CAGE Challenge, a public competitionto build an autonomous network defender agent in a high-fidelity networksimulator. Our approaches all build on the Proximal Policy Optimization (PPO)family of algorithms, and include hierarchical RL, action masking, customtraining, and ensemble RL. We find that the ensemble RL technique performsstrongest, outperforming our other models and taking second place in thecompetition. To understand applicability to real environments we evaluate eachmethod's ability to generalize to unseen networks and against an unknown attackstrategy. In unseen environments, all of our approaches perform worse, withdegradation varied based on the type of environmental change. Against anunknown attacker strategy, we found that our models had reduced overallperformance even though the new strategy was less efficient than the ones ourmodels trained on. Together, these results highlight promising researchdirections for autonomous network defense in the real world.

Beyond CAGE: Investigating Generalization of Learned Autonomous Network Defense Policies