Previous Projects
Project Number: NPRP 09 - 256- 1-046
Cycle: 3
PIs: Dr. Chris Clifton
Co-PI: Prof. Qutaibah Malluhi
Duration: Nov 2010 - Oct 2013
Funding agency: QNRF
Collaborating institutes: Purdue University
Abstract:
Cloud computing, and in particular outsourcing of data management, is of growing interest. The benefits of sharing computing resources, and co-locating computing resources and energy sources, include reduced personnel cost, sharing cost of peak load provisioning, etc. Outsourcing data management can also enable intellectual value added to clients from analysis of their data. Significant knowledge can be gained from aggregation and analysis of data from a variety of sources. Unfortunately, this also increases potential for misuse of data, raising privacy and security concerns. Privacy regulations on trans-border transport of data can prevent such outsourcing. This project will develop techniques and a proof-of-concept prototype for managing data in the cloud that encrypts the link between identifying information and sensitive values, ensuring that individually identifiable sensitive information is only known to the client. As most privacy laws apply only to individually identifiable information, this leaves the cloud datacenter free of privacy regulations and prevents misuse of the data to harm the individual. This could significantly expand the trans-border market for cloud datacenters. Key technical challenges to be addressed include parameterizing anonymity metrics to meet legal standards, query processing techniques supporting partially encrypted data, analysis techniques for data partially encrypted to meet privacy rules, and distributed and parallel processing.
Project Number: NPRP 4- 1454 - 1- 233
Cycle: 4
PIs: Prof. Qutaibah Malluhi Co-PI: Dr. Umit Catalyurek
Duration: Apr 2012 - Mar 2015
Funding agency: QNRF
Collaborating institutes: Nile University and Ohio State University
Abstract:
In recent years, two revolutionary paradigms have emerged: one in life science and one in computer science. In life science, the newly available Next Generation Sequencing (NGS) technology is providing low-cost high throughput methods for determining the nucleotide sequence of whole genomes. In computer science, the cloud computing paradigm is providing means by which computational resources needed for compute intensive jobs can be acquired on-demand on pay-per-use basis. However, current cloud systems can effectively support neither management and process of large datasets nor parallel applications that require high-bandwidth/low-latency networks or emerging accelerator architectures. In this project, we tackle this problem, and design and develop necessary cloud broker infrastructure, as well as efficient high performance parallelized algorithms for analysis of emerging NGS datasets to develop a complete solution for on-demand genome sequence analytics
Project Number: NPRP 04- 1534- 1- 247
Cycle: 4
PIs: Prof. Qutaibah Malluhi
Co-PI: Dr. Walid Aref
Duration: Apr 2012 - Mar 2015
Funding agency: QNRF
Collaborating institutes: Purdue University
Abstract:
Dealing with large data is no longer the exclusive domain of big labs. Recent technological innovations have greatly increased the rate at which scientific data is collected, and have made scientific data easily accessible to small teams of scientists. However, the cost of storing, analyzing, and sharing data, as well as maintaining the needed infrastructure are too high for small labs to bear. Cloud computing comes handy as it reliefs small labs from the software and hardware maintenance as well as provides massive computing and storage capabilities as needed without the associated overheads. We propose to develop and deploy a cloud-enabled administration-free scientific data manager and collaboration environment. This will offer scientific data management and analysis tools via easy-to-use cloud-based data and analysis workspaces that will enable scientists to tackle large scientific problems with the least IT overhead. The proposed system innovatively tackles key requirements of managing scientific data including provenance and annotation management, supporting dependencies involving user actions, and similarity-based query processing. This system will facilitate collaborations among scientists while guaranteeing the right levels of security and privacy of the scientists' data.
Project Number: NPRP 09- 622-1-090
Cycle: 3
PIs: Prof. Qutaibah Malluhi
Co-PI:
Duration: Nov 2010 - Oct 2013
Funding agency: QNRF
Collaborating institutes: Purdue University
Abstract:
Cloud computing has grown to be one of the fastest growing segments of the IT industry. In such open distributed computing environments, security is of paramount concern. This project aims at developing techniques and tools for private and reliable outsourcing of compute-intensive tasks on cloud computing infrastructures. The project enables clients with limited processing capabilities to use the dynamic, costeffective and powerful cloud computing resources, while having guarantees that their confidential data, and the results of their computations, will not be compromised by untrusted cloud service providers. Moreover, the proposed methods enable carrying out cloud computations in a cheating-resilient manner, where the client can detect incorrect cloud computation answers with high probability. In the proposed techniques, the client would only do work which is linear in the size of its inputs, and the cloud bears all of the super-linear computational burden. Moreover, the cloud computational burden would have the same time complexity as the best known solution to the problem being outsourced. This prevents achieving secure outsourcing by placing a huge additional overhead on the cloud servers. These proposed techniques are even useful with trusted cloud servers, providing a "defense in depth" where any damage resulting from the compromise of a remote server is confined to that server's data and does not extend to the clients that use it.
