In the 20 century the cloud computing is emerged frequently. The cloud computing is the meshed network through which data can be shared easily. Cloud computing is considered as service oriented. The term “Cloud computing” was introduced when these services and application were brought up to access of internet. These services and application can easily be used on devices and gadgets as well as there is no need of any special application to run these type of services and application. There are many known companies that are providing services from the cloud. “Google” has a private cloud that delivers Google docs, document applications, web analytics tools, maps and email access etc. “Microsoft” a multinational technology company that sells computer software is also providing the cloud services for content and business intelligence tools as well as “Office 365” that is available on internet for users. “SalesForce.com” Runs its application set for its clients in a cloud, and its Force.com and Vmforce.com merchandise furnish builders with structures to construct customized cloud services. The cloud computing is also divided in different types of cloud like Public Cloud, Private Cloud, Hybrid Cloud, Community Cloud. Cloud computing has won big attention in the scientific community. Cloud computing is a mannequin for enabling convenient, on-demand community get entry to a shared pool of configurable computing resources that can be unexpectedly provisioned and released with minimal management effort (Rebollo, 2012). This definition describes cloud computing as having five characteristics i.e., on-demand self-service, wide network access, resource pooling, fast elasticity, and measured service. Although there are many advantages of adopting cloud computing, there are additionally good sized obstacles for adoption. One of the most extensive barrier to adopt cloud is security (El-Gazzar, 2014). As cloud computing represents an extraordinarily new computing paradigm, consequently the most necessary subject is its security from each the standpoint of cloud consumer and Cloud Service Provider (CSP). Migrating critical functions and crucial data to the cloud surroundings is of super concern for agencies that are moving beyond their record centers. To mitigate these concerns, a CSP must ensure that customers will continue to have the identical protection and privacy controls over their functions and services and supply evidence to clients that their organization are impervious and they can meet their provider degree agreements (Rittinghouse, 2009). Since the emergence of cloud computing in 2006, a lot of assessment papers based on cloud computing are on hand in the modern literature but to date, no systematic evaluate of cloud computing issues has been published. Therefore, the main intention of this lookup is to systematically pick out and evaluate posted research work and provide an overview of risk analysis, security techniques, threat severity, and effect of these problems on cloud customers and providers.
2 Research Questions
What are the risks associated with cloud computing from cloud customer’s perspective?
What are the approaches introduced and how the approaches have been validated to ensure data security in cloud computing?
What are the data security techniques used for protecting and securing data in cloud throughout the world?
3 Related Work
Cloud is composed on the architecture of Service Models and Deployed Models.
3.1 Service Model:
I. Software as a Service (SaaS)
It is the top layer of cloud service model. The cloud service provider developed and hosts the software program or software on the cloud infrastructure permitting the customers to use it with a variety of gadgets via the usage of the optimized consumer interface such as net browser. However the underlying cloud infrastructure, network, servers, running structures or even single software abilities is no longer manageable by the customers (BalaNarayada, 2013). It helps the customers to save his future expenses because of licensing of the usual packages is extra costly compared to the month-to-month rate for renting the software from cloud service.
II. Platform as a Service (PaaS)
A center layer of cloud service model that provides a software surroundings or platform for the users to design, develop, install and check their application besides annoying about the underlying of the cloud infrastructure the usage of the virtual servers of the cloud service provided (BalaNarayada, 2013). Therefore, the customers can build their personal applications which run on the provider infrastructure and they have manipulate over the deployed application they built.
III. Infrastructure as a Service (IaaS)
The person allowed to rent the processing, storage and other fundamental computing assets to deploy and run arbitrary software program which consists of operating machine and functions and they have manipulate over the running gadget and network. It provides basic storage and computing capabilities. It also has a facts Centre area that can assist to handle workload (BalaNarayada, 2013).
3.2 Deployment model
1. Public Cloud
The entire infrastructure of this cloud model is placed on the premises of the cloud carrier provider. The users usually share the identical infrastructure pool with constrained configuration. It is accessible via any consumer and any user can get their information in the same cloud supplied with the aid of the cloud service provider. It offers scalable, dynamically provisioned and virtualized sources handy over the Internet.
2. Private Cloud
The cloud infrastructure is owned by means of only one person and it is not shared with the others. The user has physical manage over the cloud infrastructure and it is more impervious in contrast to the public cloud the place every person shares the same cloud infrastructure. It presents host services on the non-public network that helps most company community and records administrators to emerge as in-house service company successfully. In (R, et al., 2011) it was elaborated a perception of a private cloud that addresses the necessities and desires of e-learning and collaboration in university.
3. Hybrid Cloud
Combination of the public, the private or even the community cloud infrastructure which allowed the transitive data exchange. It improves the flexibility of the cloud infrastructure the place where the customers can put into effect to the personal cloud and the usage of the public cloud resources.
4. Community Cloud
The cloud infrastructure is shared amongst companies that share the same issues such as the mission, protection requirement, and policy. It may additionally be owned through greater employer and it can exist on premises or even off-premises.
Figure 1: Deployment Models of Cloud (Cloud Security Alliance, 2013)
A brief study was done upon the risk analysis, security issues and techniques to secure the cloud computing. First discussing about the risk analysis it’s divided in different categories associated with cloud computing from customer perspective.
4.1 RQ1: The risks associated with cloud computing from Cloud Customer’s Perspective
a) Data Security, Privacy ; Control Risks
• User Access:
The purchaser is completely accountable for the management of all software program safety controls. These consist of software access control, software patching, and viruses’ protection (Julisch & M, 2010). One of the dangers is how a customer face the privileged fame of CSP and security problems such as fault elimination, data harm and data migration.
• Data Privacy and Security:
It is an indispensable security concern for the end- customers to recognize about the privacy and protection of their records from CSP in order to make sure that data privations is no longer compromised. But eventually, the customers are accountable for the safety and integrity of their personal data even it resides on provider’s premises. The loss of encryption key or privileged get entry to code will deliver a serious trouble to cloud provider users. Accordingly, lack of cryptographic administration information will heavily lead crucial damages of information loss, and unexpected leakage of consumer facts to the outside world. Customer information and business secrets ought to now not be leaked while residing on CSP premises. According to the CSA team (CSA, 2013), the burden of fending off information loss does no longer fall totally on the provider’s shoulder. If a client encrypts information before setting it to the cloud and misplaced the encryption key, the records will be misplaced as well.
• Data Segregation:
It is the responsibility of cloud purchaser to discover out the techniques used by the provider to segregate the data and have to make sure that the encryption schemes are deployed and are advantageous adequate to supply security. Encryption can’t be assumed as the single answer for data segregation problem. In some cases, clients may also no longer prefer to encrypt statistics because encryption accident can smash the data. (Rosado, 2012)
• Data Availability:
Customer’s private data and statistics on the Cloud is no longer in hand both misplaced or hack, it is hard to retrieve the original data.
• Secure Data Deletion
Appropriate, error-free and well-timed records deletion can also be impossible and undesirable. From two, One of the reasons is the more copies of records reside at extraordinary places and the second is that the disk to be destroyed also carries information from other clients. Data is supposed to be destroyed totally when it is no longer required. But due to the physical traits of the storage medium, the records deleted may also nevertheless exist and can be restored. This may purpose a hazard of sensitive information disclosure to the client.
b) Technical Risks
• Infrastructure Capabilities:
It is challenging to exhibit CSP that their cloud performance is no longer in accordance with their agreed SLA due to the fact of the server’s workload and variable nature of the network. This purpose disputes and litigation. The solution is to consider the cloud performance under suitable investigation before adopting. Another answer is to use third-party monitoring equipment for the verification of system performance.
• Application Development:
The reason is to permit developers to enhance their applications over the supplied platform. Therefore, the clients are mainly responsible for defending their developed purposes and the platform. At the same time, the providers are responsible for separating the customer’s applications and improvement environments.
According to (Popovi?, 2010), the threat of compatibility arises if the client wishes to go from one provider to the other due to the fact the storage offerings supplied through one CSP might also be incompatible with every other provider’s service.
c) Compliance and Audit
• Disaster Recovery:
Cloud Customer need to be aware of what will manifest to their facts if a catastrophe occurs. Therefore, it is the customer’s essential security duty to ask whether or not the provider will be able to totally recover your data and how long it will take.
• Legal challenges:
CSP is greater inclined to legal and regulatory concerns and commits to preserve and process customers’ data in particular jurisdictions that supply protection and privacy of information as promised in their SLA’s. Even then, the corporations are mostly accountable for the privations of their data saved at the CSP website online. The pc processing strength or storage one buys through a Cloud service may additionally be primarily based in another country or may be divided between a couple of countries. Despite the benefits of value and efficiency, it raises legal problems through exporting customer’s data abroad (Popovi?, 2010).
d) Physical Security
• Data Location:
As the information is saved redundantly in a couple of physical locations via the CSP and that vicinity statistics is no longer published to the customer. On the client side, it is tough to determine whether or not fantastic safety measures are in region to tightly closed customers’ data. The purchaser cannot avoid the downtime of a cloud computing environment, which is the time in which the CSP machines are not working properly. This scenario brings big discourage to the confidence of customers.
4.2 Approaches that’s been introduced to secure cloud computing
List of Approaches to ensure Security of Cloud Computing Encryption
Data Concealment Component
Table 1: Approaches to Ensure Data Security in Cloud
Figure 2: Approaches usage (Irfan Khan ; , 2014)
According to research there are eight approaches that are been proposed to ensure the data security in cloud computing. The analysis above shows that most common approach is Encryption of data. A more brief literature review will be done on each of approaches.
The outcomes show that most common method was once encryption (45%) to assure the statistics security in cloud. In (Somani, 2010) a digital signature with RSA, algorithm scheme is proposed to make sure the facts protection in the cloud. In which software program used to crunch down the information files into few lines by means of the use of “hashing algorithm”. These traces are referred to as message digest then software encrypts the message digest with his private key to produce the digital signature. Digital signature will be decrypted into message digest by the software with own non-public key and public key of sender.
In (Shuai, 2011) RSA algorithm used to encrypt the records and Bilinear Diffie-Hellman to ensure the safety whilst replacing the keys. In the proposed technique a message header is delivered in the front of each records packet for direct and protected communication between client and cloud besides any 1/3 celebration server. When a consumer sends the request to the cloud server for information storage then cloud server creates the consumer public key, personal key and user identification in a certain server. Two duties performed at consumer quit earlier than sending the file to the cloud, first add message header to the data and secondly encrypt information consisting of message header by means of using a secret key. When a consumer request for data to the cloud server then it will check the message header of received data and pick up the Unique Identification for Server in cloud (SID) information. If SID data is located it will reply the person request otherwise request will be discarded.
In (sriram, 2011)Playfair and vigenere cipher, strategies had been merged with structural components of Simplified Data Encryption Standard (SDES) and Data Encryption Standard (DES). In which 64 bit block size of plain text is taken which is constant and this sixty four bit plaintext is divided into two halves by means of the usage of the “black box” the right half of has 2 bits whereas left 1/2 has 6 bits, then these 6 bits are feed into “superior function” block where these 6 bits are in addition separated in two halves where first two bits characterize the rows and final 4 bits symbolize the column with the aid of figuring out the rows and column the corresponding fee can be selected. Then this characteristic is applied to all 8 octets of the output of vigenere block the resultant of the black box is a acquire of sixty four bits then these bits are in addition divided into 4 new octants in a similar way proper four bits are unified to formulate proper halves. Finally left and proper halves are XOR-ed to obtain left 1/2 of this arrangement. This manner is repeated three times.