Abstract
Modern observability systems treat logs, traces, and metrics as noisy data sources requiring denoising, aggregation, and statistical anomaly detection. However, this assumption breaks down in adversarial, highly distributed, or rapidly evolving systems where signals may be incomplete, obfuscated, or implicitly deceptive due to system complexity, emergent behaviors, or malicious interference.I propose the Stratified Deception-Inference Architecture (SDIA), a hierarchical cognitive framework that reinterprets observability data as an adversarial communication system rather than a passive telemetry stream. Inspired by strategic deception paradigms from classical warfare, SDIA introduces a multi-layer architecture that explicitly models noise, semantic drift, and hidden intent as first-class computational objects.
SDIA reframes anomaly detection as latent intent reconstruction under uncertainty, enabling systems to infer failure modes, attack strategies, and emergent misconfigurations through adversarial reasoning, temporal semantic modeling, and multi-agent inference.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Manuel-Devadoss, Johny, "Stratified Deception-Inference Architecture (SDIA) - A Multi-Layer Adversarial Intelligence Framework for Log Semantics, Intent Inference, and Deception-Aware Observability Systems", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/9978