ABSTRACT:

Single cloud management platform technology has reached maturity and is quite successful for information technology applications. Enterprises and application service providers are increasingly adopting a multi-cloud strategy to reduce the risk of cloud service provider lock-in and cloud black-outs and, at the same time, get the benefits like competitive pricing, the flexibility of resource provisioning and better points of presence. Another class of applications that are getting cloud service providers increasingly interested in is the carriers' virtualized network services. However, virtualized carrier services require high levels of availability and performance and impose stringent requirements on cloud services. They necessitate the use of multi-cloud management and innovative techniques for placement and performance management. We consider two classes of distributed applications - the virtual network services and the next generation of healthcare - that would benefit immensely from deployment over multiple clouds. This thesis deals with the design and development of new processes and algorithms to enable these classes of applications. We have evolved a method for optimization of multi-cloud platforms that will pave the way for obtaining optimized placement for both classes of services. The approach that we have followed for placement itself is predictive cost optimized latency controlled virtual resource placement for both types of applications. To improve the availability of virtual network services, we have made innovative use of the machine and deep learning for developing a framework for fault detection and localization. Finally, to secure patient data flowing through the wide expanse of sensors, cloud hierarchy, virtualized network, and visualization domain, we have evolved hierarchical autoencoder models for data in motion between the IoT domain and the multi-cloud domain and within the multi-cloud hierarchy.

Complete Thesis in Adobe Acrobat format.