Reconsidering big data security and privacy in cloud and mobile cloud systems

Large scale distributed systems in particular cloud and mobile cloud deployments provide great services improving people's quality of life and organizational efficiency. In order to match the performance needs, cloud computing engages with the perils of peer-to-peer (P2P) computing and brings up the P2P cloud systems as an extension for federated cloud. Having a decentralized architecture built on independent nodes and resources without any specific central control and monitoring, these cloud deployments are able to handle resource provisioning at a very low cost. Hence, we see a vast amount of mobile applications and services that are ready to scale to billions of mobile devices painlessly. Among these, data driven applications are the most successful ones in terms of popularity or monetization. However, data rich applications expose other problems to consider including storage, big data processing and also the crucial task of protecting private or sensitive information. In this work, first, we go through the existing layered cloud architectures and present a solution addressing the big data storage. Secondly, we explore the use of P2P Cloud System (P2PCS) for big data processing and analytics. Thirdly, we propose an efficient hybrid mobile cloud computing model based on cloudlets concept and we apply this model to health care systems as a case study. Then, the model is simulated using Mobile Cloud Computing Simulator (MCCSIM). According to the experimental power and delay results, the hybrid cloud model performs up to 75% better when compared to the traditional cloud models. Lastly, we enhance our proposals by presenting and analyzing security and privacy countermeasures against possible attacks.