Blockchain-Enhanced Secure Storage for Hadoop-Based Big Data Platforms in Cloud Environments

Abstract

The high rate of growth of large data in cloud-based systems requires secure, reliable, and efficient distributed storage solutions. Although Hadoop remains a platform for scalable data management, its centralized metadata, susceptibility to quantum attacks, and outdated encryption standards are major weaknesses. This paper proposes a high-security Hadoop storage architecture to resolve these problems. We use Raft consensus to deal with decentralized metadata and Merkle trees to verify integrity. To guarantee security, the hybrid encryption algorithm based on dual-thread Elliptic Curve Cryptography (DH-ECC), a combination of quantum-resistant algorithms Kyber and Falcon, is used to provide forward-secure security. Experimental evidence shows significant enhancements compared to conventional Hadoop deployments using ZooKeeper. The proposed model decreases metadata failover latency from 210 ms to 147 ms, a 17.8% reduction. Moreover, a pipeline with GPU-accelerated encryption achieves an encryption throughput of 0.81 GB/s, 27 percent higher than the AES implementations on a CPU. DH-ECC has a speed of 4.6 GB/s in workload analysis, which is higher than AES (69) and DES (109). With RS (10,6) erasure coding (17.8% storage overhead), a significant improvement over 381/3 triple replication, and Merkle-tree verification in less than 10 ms. Quantum-resistant Kyber encapsulation incurs a negligible additional computational cost of 1.5 ms. This fact demonstrates the model's ability to provide scalable, enterprise-level, and quantum-resilient cloud storage for high-stakes industries such as finance and healthcare.