For verification of this paper, please visit on http://www.socialresearchfoundation.com/anthology.php#8

Big data is a term which emerged to take care of huge volumes of data which involves new architectures and technologies. Big Data contain huge amount, simple or complex, increasing data groups from multiple, various sources. With the rapid development of data storage, and the data collection capacity, Big Data are now growing expanding in all streams especially in science and advance tools and technologies. Large amount of data travels from one part to another and lots of data is processed in various day to day activities. The urgent need to manage and find cures for covid -19 pandemic has made it necessary to share information in large amount between national and international organizations. The sharing of information will lead to the transfer of big data, which certainly will overstep on the privacy of an individual. Big data implementation has changed the face of traditional technologies and also introduced new security models and approaches to deal with upcoming security challenges. In this paper we discusses about the security challenges and issues related to traditional security spheres and also provide explanation that how big data overcame these issues. It will also focuses on basic study of various security aspects and issues which arises because of big data implementation and suggestions to overcome these issues. Finally, it also introduces various practices and techniques for providing security of big data in different use cases and implementation areas.

In the g Big Data there is a set of risk areas that need to be considered. These include the origin, ownership and classification of data, the data creation and collection process, and specially the lack of security procedures. Finally, the Big Data security objectives are not much different from any other data typesuesd to preserve its confidentiality, integrity and availability. In today’s era concept of Big Data is very important and complex, hence it is usual that enormous security and privacy challenges will arise (Michael & Miller, 2013; Tankard, 2012). Big Data has definite characteristics that affect the security of information. On the security solutions design that are required to handle all these characteristics and requirements, there is a direct impact of these challenges (Demchenko, Ngo, Laat, Membrey, &Gordijenko, 2014). Currently, no complete security solution isexisting.

Implecations of Study:- Traditional security solutions are mainly prepared to protect small amounts of static data. They are not sufficient to the satisfy the requirements imposed by Big Data services Hence there is no single mystic solution to solve the security and privacy challenges of Big Data. (Cloud Security Alliance, 2013). There is a requirement to understand that how the collection of large amounts of complex unstructured  and structured data can be protected. The Simple and more common solution for this can be encryption i.e. encrypting everything to make data secure regardless where the data resides on data center, mobile devices, on   computer or any other place. As  long as Big Data grows and its processing gets faster, then encryption, tokenization and masking can work as critical elements to protect important and  sensitive data. (Tankard, 2012).It is important for Big Data projects  to take into consideration the identification of the various data sources, the source and originators of data, as well as the data access  rights. It is also important to make a proper classification to identify critical data, and align with the security policy of the organization information in terms of access control and data handling policies. (Kindervag, Balaouras, Hill, &Mak, 2012).Now a days, the big Data security solutions spread the secure perimeter from the private enterprise to the public cloud (Juels & Oprea, 2013). In this way,across domains, a trustful data provenance mechanism should be also created. In addition, similar mechanisms (Luo, Lin, Zhang, & Zukerman, 2013) can be used to mitigate distributed denial-of-service (DDoS) attacks launched against Big Data infrastructures. Also,throughout the entire data lifecycle – from data collection to usage, a Big Data security and privacy is necessary. A recent work describes given privacy extensions to UML to quickly visualize privacy requirements to help software engineers, and design them into Big Data applications (Jutla, Bodorik, & Ali, 2013). While implementing Big Data security it becomes compulsory that mechanisms that report legal requirements about data handling, need to be met. Secure encryption technology should be working to protect and secure all the confidential data.  In order to be successful, These mechanisms need to be transparent to the end-user and have less effects on the performance and scalability of data (Advantech, 2013). Fully Homomorphic Encryption (FHE) (Gentry, 2009), Secure Function Evaluation (SFE) (Lindell&Pinkas, 2002) and Functional Encryption (FE) (Goldwasser et al., 2014), and partition of data on non-communicating data centers, can help  insolving the limitations of traditional security techniques. Homomorphic encryption is a type of encryption that can  allow a specific types of computations (e.g. RSA public key encryption algorithm) to be applied on ciphertext and produce an encrypted result which, , matches the result of operations performed on the plaintext, when decrypted (Gentry, 2010). Fully homomorphic encryption has various applications (Van Dijk, Gentry, Halevi, &Vaikuntanathan, 2010). It  allows encrypted queries on databases, which stores secret and  private information of user where that data is normally stored.(Ra Popa& Redfield, 2011). It also enables private queries to a search engine.  When the user submits an encrypted query and the search engine generate a brief encrypted answer without  looking at the query which could contain private information of the user. The homomorphic encryption method also enables the  searching on encrypted data. When a user stores encrypted files on a remote file server, the server cannot decrypt the files on its own and  later the server retrieve  files only when it satisfy some boolean constraint. Generally, the complete homomorphic encryption  technique   improves the efficiency of secure multiparty computation.

Research Gap:- Now a days enterprises are implementing big data like never before, using powerful analytics to drive decision-making, boost performance and identify opportunities. But a complete set of big data security concerns comes with the enormous enhancement in data usage and consumption. Finally, big data adoption comes down to one question for many enterprises: that how can you leverage big data’s potential while effectively mitigating big data security risks? The volume of data collected, stored, and processed is increasing everyday with the proliferation of devices connected to the Internet and connected to each other, which also brings new challenges in the form of information security. Although, in the Big Data infrastructure, the currently used security mechanisms such as firewalls etc cannot be used because the security mechanisms should be extended along the perimeter of the network  of the organization to fulfill the user or data mobility requirements. After consideration of these new scenarios, the important question is  that “what security and privacy policies and technologies are more adequate to fulfill the current Big Data privacy and security demands?(Cloud Security Alliance, 2013). These challenges may be organized into four Big Data aspects such as infrastructure security (e.g. secure distributed computations using Map Reduce), data privacy (e.g. data mining that preserves privacy/granular access), data management (e.g. secure data provenance and storage) and, integrity and reactive security (e.g. real time monitoring of anomalies and attacks).

The increased use of IOT, Internet, latest development in 5G networks and increased computing powers at edge devices make the problem worse. Furthermore, big data privacy and security solutions in the era of COVID-19 need new start to fulfill with more demanding data security laws and regulations across the world.

Fig:-Security and Privacy challenges in Big Data ecosystem (adapted from (Cloud Security Alliance, 2013))

A non-profit organization Cloud Secure Alliance (CSA), has created a Big Data Working Group that has focused on the major challenges to implement secure Big Data services (Cloud Security Alliance, 2013). CSA has considered the various security and privacy issues and challenges into four different aspects of the Big Data system. These aspects are Data Privacy, Infrastructure Security, Reactive Security, Data Management and, Integrity . Each of these aspects hasmany security challenges, as per the CSA. That can be Infrastructure Security, Secure Distributed Processing of Data, Security Based Actions for Non-Relational Data-Bases, Data Privacy, End-to-End Filtering & Validation etc These data security and privacy challenges cover the overalllifecycle of the Big Data spectrum likebase of data production , the data itself, data processing, data storage, data transport and data usage on different devices.

1. https://www.sisense.com/glossary/big-data-security/

2. Security and Privacy Issues of Big Data, José Moura1, Carlos Serrão 2, 1University ofLisboa, Portugal 2 IT, Instituto de Telecomunication, Lisboa, Portugal

3. Security Issues and Challenges related to Big Data, Sonali Vyas1, Sameer Saxena2, Amity University Rajasthan

4. http://www.cra.org/ccc/files/docs/init/bigdatawhitepaper.pdf

5. http://www.nessi-europe.com/Files/Private/NESSI_WhitePaper_BigData.pdf

6. http://sites.amd.com/sa/Documents/IDC_AMD_Big_Data_Whitepaper.pdf

7. Kogge, P.M.,(20-24 May,2013), “Big data, deep data, and the effect of system architectures on performance”

8. Szczuka, Marcin,(24-28 June,2013),” How deep data becomes big data”.

9. Sharma PP, Navdeti CP. Securing big data hadoop: a review of security issues, threats and solution. Int. J. Comput. Sci. Inf. Technol. 2014; 5 (2):2126-31.

10. Richards NM, King JH. Three paradoxes of big data. Stan. L. Rev. Online. 2013; 66: 41.

11. Bharati, T. S. (2015). Enhanced Intrusion Detection System for Mobile Adhoc Networks using Mobile Agents with no Manager. International Journal of Computer Applications, 111(10).

12. Bharati, T. S., & Kumar, R. (2015, March). Secure intrusion detection system for mobile adhoc networks. In Computing for Sustainable Global Development (INDIACom), 2015 2nd International Conference on (pp. 1257-1261). IEEE.

13. Bharati T. S. (2017). Agents to Secure MANETS. International Journal of Advanced Engineering and Research Development, 4(11), 1267-1273.

14. Jaseena KU, David JM. Issues, challenges, and solutions: big data mining. CS & IT-CSCP. 2014 Dec 27; 4 (13):131-40. 15. Abouelmehdi K, Beni-Hessane A, Khaloufi H. Big healthcare data: preserving security and privacy. Journal of Big Data. 2018 Dec 1; 5(1):1.

16. Li N, Li T, Venkatasubramanian S. t-closeness: Privacy beyond k-anonymity and l-diversity. In2007 IEEE 23rd International Conference on Data Engineering 2007 Apr 15 (pp. 106-115). IEEE.

17. Sweeney L. Achieving k-anonymity privacy protection using generalization and suppression. International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems. 2002 Oct;10(05):571- 88.

18. Sweeney L. k-anonymity: A model for protecting privacy. International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems. 2002 Oct;10(05):557-70.

19. Samarati P. Protecting respondents identities in microdata release. IEEE transactions on Knowledge and Data Engineering. 2001