부산대학교 도서관

With the rapid growth of Big Data, securing its storage has become crucial. This study proposes to enhance the secure storage of big data at rest in Hadoop by improving encryption algorithms. The Elliptic Curve Cryptography Algorithm (ECC) is upgraded by a parallel two-threaded approach for unstructured data. For structured data, enhance Paillier Homomorphic Encryption to support operations on ciphertexts. Experiments on datasets up to 4 G show that the modified ECC method reduces encryption time to 60 - 80 seconds, compared to 100 - 160 seconds for standard ECC, AES, and DES. It can also use shorter key lengths than RSA with comparable levels of security. Enhanced Paillier encryption uses large prime numbers to ensure the validity of the ciphertext. By combining these improved encryption techniques within a secure Hadoop framework, this research demonstrates an effective way to address vulnerabilities in Big Data storage.