Trust-Aware Blockchain Security with Differential Privacy and Zero-Knowledge Proof for Resilient Digital Economy Communication Ecosystems

Abstract

With the rapid expansion of the digital economy ecosystem, blockchain has become the core technology and theoretical path to support data interaction and trusted transactions. However, it still faces significant challenges in terms of privacy protection, transaction traceability and anti-attack capabilities. To address the above challenges, this study proposes a dynamic trust-aware blockchain security algorithm (DTBCSA) based on differential privacy and zero-knowledge proof. This algorithm is used to improve the system robustness and privacy protection capabilities in multiple scenarios. The core design of the algorithm includes: (1) introducing a dynamic trust evaluation mechanism. This mechanism dynamically adjusts the trust level of the node by analyzing its behavioral characteristics and historical interactions. At the same time, the secondary authentication mechanism is activated when the risk threshold is triggered; (2) embedding a differential privacy mechanism in the computing power transaction and model training process. This mechanism protects sensitive data and model parameters through Laplace noise; (3) using zero-knowledge proof to ensure the non-repudiation of transactions. At the same time, the aggregation and parallel optimization strategies are adopted to significantly reduce the computational overhead of proof and verification. In the experimental part, DTBCSA reduced the acceptance rate of malicious transactions from 70% of the baseline to 8% in the computing power trading market simulation. At the same time, it reduced the reputation distribution imbalance (Gini) from 0.42 to 0.20. In the federated learning collaborative training experiment, DTBCSA reduced the success rate of member inference attacks from 81% to 38% under the condition of privacy budget ε=1. While maintaining privacy protection, it improved the model accuracy by about 1.8 percentage points compared to FedDP. In addition, in the zero-knowledge proof verification performance test, DTBCSA reduced the proof size and generation time through aggregation optimization.