Security Mechanisms and Privacy Issues In Cloud Storage System

The demand for cloud computing increases day by day and this also increases the need

for Security for the same. Several researchers focused on the objective to provide

security to the cloud platform but till now no one ensures it. So, similar to the same

focus firstly the study of different methods like Encryption, authentication, and many

others discussed in this work and then based on the recent analysis some of the methods

were selected in the work to propose a novel approach. This work is done in three

different phases which involved (a) Cryptography based Frameworks, (b) Proposed

Fragmentation Framework, (c) Proposed Biometric-based Framework. To analyse the

performance of all these phases, five different datasets are prepared with the

combination of different types of data files like Image, Audio, and Video. In this, the

data size varies from 100 to 300 files and the maximum size of data is 289810KB. The

performance is analysed on every phase based on different parameters depending on the

phase. In phase 1, three cloud environments are generated where different security

algorithms are added and performance is analysed based on Time, Throughput, and size

of Data after Encryption. These three environments use Diffie-Hellman, proxy reencryption,

HASBE and Blowfish, RSA with ECC algorithms, and results are

calculated for all these algorithms. After analysis, it is clear that the performance of

cloud environment-3 in phase 1 is better in terms of Time, Throughput, and size of

Data. In this environment, Keyword Encryption and Hybrid Cryptographic Algorithm

are used which works based on RSA with ECC Algorithm. Though this scheme

provides end to end security in a cloud environment yet, it can be forged. So, to enhance

the framework different researchers followed fragmentation-based techniques in which

data is first divided into small portions called fragments and then after encryption, it

will be uploaded on different cloud servers to protect data against various forgeries.

Different researchers also performed well on the same but none of them focus on the

data loss due to fragmentation. As data is of different types so, the same fragmentation

will be one cause of the data loss. So, to cope up with this issue a new FragSecure

framework is proposed in the second phase of this work. In this framework, the type of

data uploaded on the cloud server plays its major role and use for fragmentation