@IJRTER-2017, All Rights Reserved 261 Provably Secure Key-Aggregate Cryptosystems with Broadcast Aggregate Keys for Online Data Sharing on the Cloud Dr.K.Srinivas Babu *1 N.Sindhu *2 Dr.K.Rameshwaraiah *3 *1,3 Professor Department of Computer Science and Engineering *2 M.Tech AbstractOnline data sharing for increased productivity and efficiency is one of the primary requirements today for any organization. The advent of cloud computing has pushed the limits of sharing across geographical boundaries, and has enabled a multitude of users to contribute and collaborate on shared data. However, protecting online data is critical to the success of the cloud, which leads to the requirement of efficient and secure cryptographic schemes for the same. Data owners would ideally want to store their data/files online in an encrypted manner, and delegate decryption rights for some of these to users, while retaining the power to revoke access at any point of time. An efficient solution in this regard would be one that allows users to decrypt multiple classes of data using a single key of constant size that can be efficiently broadcast to multiple users. Chu et al. proposed a key aggregate cryptosystem (KAC) in 2014 to address this problem, albeit without formal proofs of security. In this paper, we propose CPA and CCA secure KAC constructions that are efficiently implementable using elliptic curves and are suitable for implementation on cloud based data sharing environments. We lay special focus on how the standalone KAC scheme can be efficiently combined with broadcast encryption to cater to m data users and m0 data owners while reducing the reducing the secure channel requirement from O(mm0 ) in the standalone case to O(m + m0 ). KeywordsCloud Computing, Data Sharing, Data Security, Key-Aggregate Cryptosystem, Provable Security, Scalability, Broadcast Encryption, Semantic Security, CCA Security. I. INTRODUCTION The recent advent of cloud computing has pushed the limits of data sharing capabilities for numerous applications that transcend geographical boundaries and involve millions of users. Governments and corporations today treat data sharing as a vital tool for enhanced productivity. Cloud computing has revolutionized education, healthcare and social networking. Perhaps the most exciting use case for cloud computing is its ability to allow multiple users across the globe share and exchange data, while saving the pangs of manual data exchanges, and avoiding the creation of redundant or out-of-date documents. Social networking sites have used the cloud to create a more connected world where people can share a variety of data including text and multimedia. Collaborative tools commonly supported by cloud platforms and are extremely popular since they lead to improved productivity and synchronization of effort. The impact of cloud computing has also pervaded the sphere of healthcare, with smartphone applications that allow re- mote monitoring and even diagnosis of patients. In short, cloud computing is changing various aspects of our lives in unprecedented ways. Despite all its advantages, the cloud is susceptible to privacy and security attacks, that are a major hindrance to its wholesome acceptance as the primary means of data sharing in todays world. According to a survey carried out by IDC Enterprise Panel in August 2008 [1], Cloud users regarded security as the top challenge with 75% of surveyed users worried about their critical business and IT systems being vulnerable to attack. While security threats from external agents are widespread, malicious service providers must also be taken into consideration. Since online data almost always resides in shared environments (for instance, multi- ple virtual machines running on the same physical device), ensuring security and privacy on the cloud is a non trivial task. When talking about security and privacy of data in the cloud, it is important to lay down the requirements that a data sharing service must provide in order to be considered secure. 1.1 The Key-Aggregate Encryption Scheme The most efficient proposition pertaining to our problem statement, to the best of our knowledge, is made in [6]. The proposition is to allow Alice to combine the decryption power of multiple data classes into a single key of constant size. Thus, while each class of data is encrypted using a different public key, a single decryption key of