Privacy and Data Level Security Implementation in Cloud Environments

Abstract

Cloud Computing is one of the main emerging technologies in the world that supports to use services that can be carried out on behalf of users on hardware that the users do not own or operate. But many people yet find it difficult to get adapted to using cloud applications due to security risks and failures. There is a threat of losing or leaking sensitive data or documents. Therefore one major problem in cloud services and applications is building clouds and its applications to protect the user’s privacy. So as a solution to the above issue, it has been planned to implement a cloud application system added with a proper client based privacy management component that will ensure the user that even the cloud service provider will not be able to view the consumer data. In Sri Lanka there are no local cloud services available for local companies. Therefore they have to use external clouds which make more risk. This will again limit the usage of cloud services due to cost, compatibility and availability factors. So the proposed system will be targeting the local organizations to provide services to store and manage their sales and customer data added with the privacy management component. In solving the problem, the proposed solution is to build a proof of concept cloud application to manage sales and customer data for Sri Lankan organizations, with the privacy management feature. The solution is based on the concepts of grid computing, data and task parallelism, map reduce and data encryption. The user data is entered to the cloud after going through an obfuscation process. The data is broken down into chunks and then distributed across a grid and processed in parallel nodes enhancing the security of the application. The output is retuned back to the user using map reduce operations. A Linux/Windows based cloud platform will be used to host the cloud application. Grid Gain Open source infrastructure will be used to implement the cloud platform. Windows Server edition will be used to host the cloud platform. To implement the Cloud based application Java/ASP.net will be used. SSL is implemented for the communication channel between the server and the clients. Finally an evaluation is done on how the proposed solution will work on large scale clouds using the CloudSim simulation platform. To achieve this I have successfully implemented the map reduce logic in CloudSim simulation platform.