No abstract

The proliferation in the popularity of the cloud-based data storage services motivated the data owners to store a huge amount of confidential files on the remote servers in an encrypted format. The users/clients can send their queries to the database owner to retrieve the data files in the encrypted database while protecting privacy of both the queries and the database. The database owners can outsource their enormous biometric data and identification tasks to the cloud server such as Amazon to avoid high storage and computation costs. However, this adds potential threats to the privacy of users. This paper presents a combined Biometric-based Bucket Encrypting Index Structure with Random Generator (B2EIS-RG) for efficient and privacy-preserving biometric identification outsourcing in the cloud. The encryption process includes multi-keyword query processing along with the conjunction and disjunction logic queries to ensure high privacy guarantee against the keyword attacks. Experimental evaluation over a large dataset demonstrates that the proposed scheme can achieve modest time efficiency, and they are practical for use in the huge encrypted database systems. The proposed scheme is found to be highly secure even if the attackers can forge the biometric identification requests.

Certificateless Public Key Cryptography (CL-PKC) scheme is a new standard that combines Identity (ID)-based cryptography and tradi- tional PKC. It yields better security than the ID-based cryptography scheme without requiring digital certificates. In the CL-PKC scheme, as the Key Generation Center (KGC) generates a public key using a partial secret key, the need for authenticating the public key by a trusted third party is avoided. Due to the lack of authentication, the public key associated with the private key of a user may be replaced by anyone. Therefore, the ciphertext cannot be decrypted accurately. To mitigate this issue, an Enhanced Certificateless Proxy Signature (E-CLPS) is proposed to offer high security guarantee and requires minimum computational cost. In this work, the Hackman tool is used for detecting the dictionary attacks in the cloud. From the experimental analysis, it is observed that the proposed E-CLPS scheme yields better Attack Detection Rate, True Positive Rate, True Negative Rate and Minimum False Positives and False Negatives than the existing schemes.