PAST RESEARCHES ON BIG DATA SECURITY
In Big data, everyone can provide different and multiple security issues according to their opinion and the environment. However main security issues are similar to other domains that is Authentication, Authorization, Confidentiality (data), Integrity and Accounting. In addition, verifying the integrity of the data before using it for taking decision, preserving privacy of the data while processing and confidentiality of the data while transmission are other common security issues to be considered in Big data. Even though security issues in the Big data are common with some other domains, this section gives a look on security problems that have been found so far in the Big data environment by various research.
M. R. Jam et. al. (2014) discussed about the Security mechanisms of Hadoop. They have discussed about the layers of defence for Hadoop nodes. These layers are Perimeter Level Security, Authentication, Authorization, OS security and the data protection. There is a very strong security at the file system level in Hadoop, but the granular support is necessary to entirely secure access to data by users and Organization Intelligence applications, which is absent in it.
P. Adluru et. al. (2015) have proposed their own security mechanism for Hadoop. The user must be authenticated for access of distributed data and its storage on the cloud to ensure security and privacy of Big data and with the implementation of secure random encryption mechanism a more secure infrastructure can be formed. Their main goal was to present a Hadoop framework which sustains the privacy and security of the data stored on the cloud environment.
M. Kaushik and A. Jain (2014) have discussed the challenges of security and privacy in Big data. They have classified the big data security challenges into four categories such as Infrastructure Security for secure data in distributed programming environment, Data Privacy preservation in analytics and data mining with granular access control, Data Management to secure the stored data and log files with granular auditing and Integrity and Reactive Security for input validation at end-point with real-time security observations. Depending on the data and security policies of the organization, the security is applied to different security levels of data.
A. Katal et. al. (2013) have discussed about the issues, challenges and good practices of big data security. They have discussed challenges such as Privacy of big data, data access and sharing of data in secure manner along with storage and processing issues. They have also mentioned some technical challenges such as fault tolerance, scalability, maintaining quality of data and processing heterogeneous data.
Cloud Computing is used to store and process big data so they have categorized the big data security in four different categories which are network level, user authentication level, data level, and generic issues (V. Narasimha et. al., 2014). Further issues and challenges in the environment are Internode Communication,
Data Protection, Administrative rights for the clusters, Authentication of applications and nodes, and log maintenance. As a solution of these issues, Narasimha et. al. (2014) proposed layered framework for assuring cloud. They have suggested file and network encryption, logging, introduction of honeypot nodes, access control and third party secure data publication for cloud.
