Federated Learning/ FL

Definition

Federated Learning is training approach where nodes learn from local data and share updates instead of raw data. In defense applications, it improves models from field experience while reducing exposure of sensitive raw observations. The hard part is poisoned updates, non-IID data, and aggregation under intermittent links, especially when systems are deployed across contested links, coalition boundaries, and mixed human-machine teams. KhanBMS treats it as a controlled learning path for KhanBMS units operating under different theaters, tying the concept back to modular command, edge execution, and auditable authority.

Reference attributes

Layer

distributed training method

Operational value

Improves models from field experience while reducing exposure of sensitive raw observations

Primary risk

Poisoned updates, non-IID data, and aggregation under intermittent links

KhanBMS role

A controlled learning path for KhanBMS units operating under different theaters

Related terms

• Centralized Training, Decentralized Execution (CTDE)
  Training pattern where agents learn with shared global information but deploy using local observations.
• Data Poisoning
  Attack that corrupts training or fine-tuning data to implant bad behavior or degrade performance.
• Secure Model Provenance
  Cryptographic and procedural evidence tracking where a model, adapter, dataset, or artifact came from.
• Edge Model Registry
  Versioned catalog that tracks which models, adapters, signatures, and policies are deployed to tactical nodes.