A secured storage and privacy-preserving model using CRT for providing security on cloud and IoT-based applications

Abstract

Cloud computing and Internet of Things (IoT) are emerging technologies which have the capability to enhance the human daily lifestyle. Moreover, the combination of Cloud computing and IoT are increasing the productivity of the enormous number of applications in various fields like supply chains, commercial, engineering, manufacturing etc. Presently, security is a very big threat in cloud computing and the Internet of things areas. For that purpose, the various researchers have introduced many privacy-preserving models and security mechanisms in recent years. Even though, they have failed to achieve the latest cloud user's expectation in terms of security level. For satisfying the cloud users, this paper proposes a new Chinese Remainder Theorem (CRT)-based data storage mechanism for storing the user data securely in cloud database. Moreover, a new group key management scheme is also built using CRT to access the encrypted data from the cloud database. In the proposed CRT-based secured storage scheme adopts two encryption schemes which use new formulas for performing the first and second encryption and also introduced a new formula for decrypting the cloud data. In addition, a new formula is introduced in the process of group key generation for accessing the encrypted cloud data from the cloud database in a cloud server. The performance level of the security models has been evaluated by analyzing the experimental results. Finally, it is proven that the proposed data security model is better than other existing models.

Introduction

In recent years, the novelties of cloud computing techniques are growing significantly due to the rapid development of cloud users and their expectations. Resource sharing is a key feature of cloud computing that is also needed for the IoT environment. The cloud is a geographically independent platform which is able to support the cloud users for accessing the cloud data and resources from any place/any device via internet. Similarly, the IoT devices are also can access the data and resource from any location. The available cloud features like elasticity, ubiquitous access, on demand service provision and resource pooling are the basic need of the IoT technology. The efficient and versatile service is provided by elastic and on-demand feature, the reliability of service is improved by the resource pooling. These are all the convincing factors that emerges the need for combining the IoT and Cloud computing standards [43], [44].

Cloud data security is the important and essential branch of the cloud computing technology which gives protection to the cloud user's data. Here, the security policies like firewalls and virtual private networks that assure the data privacy of the cloud user. Because of the concept of resource pooling is mandatory in cloud environment the highly confidential data is also available to the third party cloud. The user authentication and secured action are necessary in cloud computing technology. In various literatures different types of security concerns and future challenges have been discussed and reviewed based on the standards like PCI-DSS, ITIL, and ISO-27001/27002. According to the changing functionalities of various architectural designs the architectural security issues also arises. Because of outsourcing the cloud computing has two main concerns: an unauthorized person can access the data in cloud without owner's permission, cloud service provider can breach the data owner's information. Any kind of security destruction will cause a massive loss and can create a critical situation in cloud. The cloud security problems should be organized, governed and strictly regulated in position for the entrepreneurs to adopt the cloud computing facilities safely.

In recent years, cloud users are increasing massively so extensive storage space and highly performing computational maintenance is needed. For the secured cloud environment, keeping the sensitive data confidentially is the very big challenge. Encryption is the only powerful mechanism that ensures data confidentiality in cloud [3]. It can protect the content and the original data is converted in to unintelligible cipher text data. When the data owner is not willing to disclose the sensitive information to the cloud databases, he/she can use encryption technique. The content-related keywords and authentication data should be embedded in the encrypted content for securing and managing data in the cloud for the administrator purpose [2]. Attribute-based encryption (ABE) is an adaptable cryptographic method for performing different functionalities. ABE is of two different types they are cipher text-policy attribute-based encryption (CP-ABE) and key-policy attribute-based encryption (KP-ABE). In cipher text-policy attribute-based encryption the cipher text have been generated by obtaining the key from the attributes set of the user. The decryption of the cipher text can be done only by using the particular secret key which is embedded in the cipher text. Where as in KP-ABE attributes and the access policy are attached in reverse order to secret keys and cipher texts [4].

An important part of cloud-based technologies is cloud storage. Day by day the user's data volume and the network bandwidth have been increasing geometrically so the capacity of the local machines cannot support the user requirements. The people were searching for a new method for storing their data, and they found more powerful storage capacity named cloud storage. Cloud storage technology is widely spreading now a day and in future storing of user's data in cloud will be a trend. Cloud storage is nothing but a computational system where the data services and the management services are provided. The devices in the cloud storage are cluster of distributed file systems, applications and network technology works together. iCloud, Dropbox, Baidu Cloud, Google Drive, etc. are the organizations which is providing cloud storage services. Because of providing substantial limit of capacity and different administrations identified with well-known applications, huge numbers of subscribers are attracted successfully by these companies. Still there are a lot of problems in security existing in cloud storage.

The significant problem among those issues is privacy problem. In the past, there were a few well-known cloud storage privacy leakage events happened. In that one is, at the year of 2014 Apples iCloud have been hacked and the personal photographs of some Hollywood actresses which are saved in the iCloud were leaked. Big scene was created by this event and the panics of the user's about the security of their information over the cloud server have been raised. The cloud service provider (CSP) manages the data in the cloud and the user has the access to store his data directly in the cloud. The physical storage of their data is not managed by the user and therefore here the division of possession and the administration of information take place. The data in the cloud is independently accessed by the CSP and there is a lot of chance for the hacker to attack the CSP server to get the clients information. As said over the two cases will cause data leakage and loss. Access restrictions or data encryption are the traditional secure storage solutions [5].

In Cloud environment, the important requirements of cloud security are access control to the user's data and web-based services. Due to the shared infrastructure and increasing number of hosting applications, unauthorized accesses to the data have been increased. New access control requirements are arising according to the environmental change. Discretionary Access Control (DAC), Mandatory Access Control (MAC), and Role-Based Access Control (RBAC) models are the traditional access control models which are identity based and identification is performed directly to the subjects. The above mentioned models are only useful for centralized systems, which are static with limited number of users with known set of services. In the internet world every user is not known to each other and domain is also not in a secure place, so the determination of access control rules that are depending over the users attributes and services will be better. Accordingly, the attribute permissions are dynamically assigned to the users and such models are named as Attribute-Based Access Control (ABAC). This type of ABAC policies are more expressive and fine grained than RBAC model but there are problems in trust, provisions for context and privacy issues [6].

The major contributions of this paper are as follows: (i) to introduce a new Chinese Remainder Theorem-based data storage mechanism to store the data securely in cloud database and (ii) to introduce a new CRT-based key management scheme which is working for group to access the encrypted cloud data from cloud database. The proposed model (iii) introduced double encryption scheme which uses new formulas. (iv) Introduce new formulas for performing encryption process in secured data storage mechanism and (v) also introduce a new formula for decrypting the encrypted data which are stored in cloud. In addition, (vi) a new formula is introduced in the procedure of group key generation to access the encrypted cloud data from cloud database in cloud server.

The remainder of this paper is arranged as follows: The more relevant and related works in the areas like encryption, decryption, CRT and cloud security have been discussed in Section 2. The overall architecture of the proposed section is given in Section 3 with proper explanation. Section 4 describe in detail about the proposed model. The experimental results are shown in Section 5 with justification for the enhancement. The conclusion of the proposed work is given in Section 6 with future enhancements.