Blockchain-Based Distributed Identity Supervision and Revocation

Decentralized Identity (DID) systems have gained wide attention due to their advantages in autonomy, interoperability, and openness. However, existing solutions still have significant shortcomings in the adaptability of malicious detection, the granularity of identity revocation, and process collaboration. This paper proposes an autonomous identity supervision and revocation scheme based on a rewritable blockchain: by introducing the XGBoost model and extracting on-chain multi-dimensional features, high detection accuracy is maintained even in extreme scenarios; meanwhile, three types of fine-grained scenarios are designed, covering voluntary deactivation by users and system-triggered malicious revocation; efficient transaction updates are achieved via chameleon hash, and compliance is ensured by combining the PBFT consensus mechanism; finally, a full-process collaborative system of "detection-revocation-evidence storage"is constructed. Experimental results show that the scheme maintains low latency in DID registration, identity verification, and identity revocation processes, operates stably, and effectively addresses the technical gaps in existing solutions.