PDDS - Improving Cloud Data Storage Security Using ... - IEEE Xplore

PDDS - Improving Cloud Data Storage Security Using Data Partitioning Technique C. Selvakumar

G. Jeeva Rathanam

M. R. Sumalatha

Department of Information Technology MIT Campus, Anna University Chennai, Tamil Nadu, Indian [email protected]



networking resources. Data are stored in the cloud through hosted network services and also it offers the use of access control in it. Cloud service provider will manage and control the cloud resources. Cloud service development encompasses services such as Software as a service (Saas), Platform as a service (Paas) and Infrastructure as a service (Iaas) and deployment models such as public cloud, private cloud and hybrid cloud. Using web browser protocols client-server works in cloud as in Fig [1]. Client uses the client devices to access a cloud system via World Wide Web.

Abstract— Cloud storage system enables storing of data in the cloud server efficiently and makes the user to work with the data without any trouble of the resources. In the existing system, the data are stored in the cloud using dynamic data operation with computation which makes the user need to make a copy for further updating and verification of the data loss. An efficient distributed storage auditing mechanism is planned which over comes the limitations in handling the data loss. In this paper the partitioning method is proposed for the data storage which avoids the local copy at the user side by using partitioning method. This method ensures high cloud storage integrity, enhanced error localization and easy identification of misbehaving server. To achieve this, remote data integrity checking concept is used to enhance the performance of cloud storage. In nature the data are dynamic in cloud; hence this work aims to store the data in reduced space with less time and computational cost.

Many independent storage servers are used in the largescale distributed storage system like cloud. The benefits of the cloud storage are flexible with reduced cost and they also manage the data loss risk and so on. Recently many work focus towards third party auditing and the remote integrity checking, providing the data dynamics. Remote archive service is responsible for properly preserving the data. The remote data integrity checking protocol detects the data corruption and misbehaving server in the cloud storage.

Keywords— Remote Data Integrity Checking, Partitioning, Error Localization, Cloud Storage.

I.

In the proposed work PDDS, remote data integrity checking is analyzed in internal and external ways. It supports data dynamics and public verification as in Fig [2] considering the untrusted server with security analysis.

INTRODUCTION

In high speed network, the Internet access becomes available in the recent years, Cloud computing is an internet based technology, being used widely nowadays to enable the end user to create and use software without worrying about the execution of the technical information from anywhere at any time . Over the network the resources are utilized and after computation these are delivered as services in cloud computing.

Fig.2. Data Storage in Cloud. Fig.1. Cloud Services Architecture

This research work aims in designing, an efficient flexible storage scheme to ensure the availability of data and data correctness in cloud, by partitioning algorithm. Data storage is

Cloud storage is a service for developers to store and access data in cloud. It deals with direct access to the storage and

c 978-1-4673-4529-3/12/$31.00 2012 IEEE

7

done by using this algorithm. Partitioning happens in vertical and horizontal directions whereby the data being used is controlled. The security mechanism is also emphasized in order to prevent unrecoverable data loss. Storage and retrieval process are simplified by reducing the storage space when there is need to store and retrieved by merging technique. II.

PROBLEM STATEMENT

A. PDDS - Design goal To ensure security and data storage efficiency in cloud, integrity checking is designed effectively. Certain phrases are being used in the content that follows. •

Dependability: Enhance the mechanisms work of the integrity checking against the service attacks and threads.

•

Lightweight: Communication and computation cost in sharing and storage of the data in cloud.

•

Error localization of data: Compute and consists fast access of the data and detect the error.

•

Storage: End user can store the data in cloud at anytime and anywhere through internet.

B. PDDS - System model Cloud storage service architecture with different network entities is represented as below. •

User: Enable end user for storing the data without any difficulty in cloud.

•

Cloud Server (CS): Manage and provide storage space, computational resources and storage services by the cloud service provider (CSP).

•

Remote data integrity checking: Integrity checking to detect and correct the data error and data localization in cloud data storage.

C. Notation and Preliminaries • F - Data files to be stored. Data in equal size and stored in block wise. • E - Encoding the files and each consists n blocks. • IN- Each individual block consists an index to represent the block when access. • FS - Data files are partitioned into pieces and stored. • Fek - Generating the public key for encoding the files. • Fdk - Generating the private key to decode the files for access. • D - Decoding the files and consists the blocks.

8

III.

RELATED WORK

In this section, literature survey is done for data integrity checking and data storage mechanisms that are currently used in dynamic multi transactional applications. The dynamic data storage with token precomputation and how it is stored in cloud is analyzed [1, 12] which provide information about effective storage mechanisms. Integrity checking concepts is also used to detect and avoid misbehaving server considering data correction and error localization. Distributed scheme is used to achieve the data quality, availability, integrity of dependable storage services. The data storage using dynamic data operation method is discussed in [2]. Security analysis is done by proxy encryption technique to encode the data. Integrity checking happens to detect the untrusted server. Distributed storage system is also used to support the forwarded data in cloud without retrieval, ensuring secured and robust data in cloud storage. Data integrity in cloud storage devices are analyzed in the research works [8, 12]. Dynamic data operation and public auditability are used for supporting the data integrity. The objective of this work is to have independent perspective and quality in services evaluating with the third party auditor. Storage model is also devised here to support multiple auditing tasks to improve efficiency. Data partitioning in vertical and horizontal directions as discussed in [6]. They partitioned data into buckets and used slicing technique for data storage. In the works [3], [4], [5], author considers generating signature methods for ensuring the cloud storage security. Dynamic operations are supported by using the RSASS method. This method discusses data integrity and data correctness stored in cloud. Reference [11] ensures remote data integrity with retrievability. Error correction and data integrity checking is used to detect the availability of data in cloud. PDP scheme with symmetric key-based cryptography for data storage security were discussed in [10]. Data availability and data error recovery mechanisms are not given much importance. In cloud storage services remote data integrity checking has many challenging issues [8], [9]. In the survey done much of the discussions are related to works, which ensures to have data copy in the local system. This limitation is overcome with the proposed approach PDDS. Token precomputation method ensures dynamic data operation and the integrity checking. This mechanism provides data storage security. The limitation with existing mechanism is, it takes more time and cost to perform the dynamic processing of data encryption and decryption techniques to store data in cloud with security. The PDDS overcomes such limitations with high performance, reduced cost and limited data storage space in cloud. It also ensures resilient against threads, attacks and misbehaving server.

2013 3rd IEEE International Advance Computing Conference (IACC)

IV.

PARTITIONING AND DOMAIN INTEGRITY CHECKING FOR DATA STORAGE

In cloud data storage system, the end users stores data in cloud and also they maintain data locally. PDDS aims in providing the integrity assurance. A. Cloud Storage Fig [3] shows how the end user is supported with dynamic data operation and security model for storing data in cloud. Unauthorized access is avoided here. It detects the threats and misbehaving server and also prevents the data from attacks. The propose data storage architecture ensures precomputation to check the corrections of the data This happens before storing the data and the dynamic data operation is done after the computation. This process enhances the security because the data are stored after the precomputation process. In precomputation the security key is generated by the encryption technique to ensure security from unauthorized access. Public and private key is generated by encryption and decryption technique to ensure security. Data integrity function is the important function in cloud storage. Normally when the data is stored the end users have to check whether the data is stored in cloud correctly or not. By the integrity checking process, the data is stored with security. Data are handled by remote data integrity checking; they do precomputation process to avoid threats. B. Access data from cloud storage service Data are retrieved from the storage service as per the end user request. As in Fig [4] the data is retrieved or restored from the server ensuring the data correction. Decryption technique processes the private key to reload the original data from the cloud server. The encoded data are decoded to view the original data without pairing the re-encryption scheme.

Fig.3. Data Storage architecture

The end user can also decide what data need to be accessed and shared by the other users in cloud. Data accessed from cloud service enables the services in secured manner. The PDDS improves performance ensuring data security during storage and retrieval of data in cloud. The data is partitioned into smaller blocks before encryption for security by generating the public key before storage. During the retrieval, the data are decrypted by generating the private key and merging the data in to the original data. Remote data integrity checking is used to maintain the data from threats. It also manages the effective storage and retrieval processes. The public auditability method manages the error localization, verification, misbehaving server and error recovery. This ensures data security from unauthorized access. It also increases the performance. Flexible access control is also provided for authentication in this work and to detect the attacks. Dynamic data operation, like insertion, deletion, and updating is also done before partitioning the data. Partitioning splits the actual data and stores in cloud. Dynamic data operation enhances cloud storage services. The data storage costs are reduced by this mechanism. C. Partitioning data Partitioning function plays an important role in this work. It splits (break up) larger files into smaller parts to store the data effectively in quick manner enhancing easy access to data also when there is need. The original data is complex and there is difficulty in storing it in cloud, so partitioning function is used to make the storage easy in cloud. The partitioned files are encrypted, that is encoded with the public key and stored in cloud. Partitioning takes place automatically when the data is fed for storing in cloud. Original file is also reconstructed when there is need to access the same. The partitioning and merging concept is provided in the following algorithm Fig [5].

Fig.4. Data access from cloud

2013 3rd IEEE International Advance Computing Conference (IACC)

9

Algorithm 1: Partitioning and Merging files 1.

Load the Input file and size.

2.

Partitioning files: Count size