ETHPrague 2026

Public blockchains expose transaction histories that enable address clustering, taint propagation, and cross-service attribution, motivating a diverse ecosystem of mixers and privacy layers. This paper presents a structured scoping review of 22 representative systems and systematizes them through a common unlinkability objective, five adversary archetypes, and an attack-surface matrix spanning chain analytics, timing inference, custodial compromise, coordination abuse, network metadata, and trusted-execution failures. The systems reviewed cover custodial Bitcoin mixers, coordinator-based protocols, Ethereum smart-contract mixers, privacy-layer protocols, privacy-native currencies, confidential-execution platforms, and exchange-mediated obfuscation. Across these families, the main finding is that nominal anonymity and cryptographic strength are poor proxies for real privacy unless operational factors such as denomination structure, cover-traffic volume, relayer behavior, and compliance interfaces are modeled explicitly. By contrasting nominal and effective anonymity and by positioning recent compliance-aware designs alongside classical mixers, we derive four design lessons for privacy-preserving blockchain systems: privacy should be pooled rather than exceptional, trust anchors should be stated explicitly, metadata leakage must be treated as part of the protocol, and compliance should be narrowed to auditable predicates rather than broad operator discretion. The resulting systematization clarifies where current systems succeed, why they fail in practice, and what next-generation blockchain privacy architectures must improve.