May 26, 2017
Jolley Hall, Room 309
"Management and Security of Multi-cloud Applications"
Adviser: Raj Jain
Single cloud management platforms like Amazon's EC2 and Microsoft Azure are common and popular today. Obtaining resources from multiple clouds would give clients competitive pricing, flexibility of resource provisioning, better points of presence and reduced risk of a total blackout. Multi-cloud management platforms that can automatically manage virtual resources on a number of clouds are an area of active research. We consider two classes of distributed applications – the virtual network services and the next generation healthcare – that would benefit immensely from deployment over multiple clouds. This research proposal deals with design and development of new processes and algorithms to enable these classes of applications. We propose a method for optimization of multi-cloud platforms that will pave the way for obtaining optimized placement for both classes of services. Placing virtual resources properly is of utmost importance from the points of view of cost and performance. The next part of our proposal is predictive cost optimized latency controlled virtual resource placement for both the types of applications. To improve availability of virtual network services we propose machine learning based fault detection and localization. Finally, to secure patient data in the wide expanse of sensors, clouds and visualization domain, our contribution will be deep learning based anomaly detection technique for data at rest and in motion.