Kumar, Pardeep

Profile Picture

Filters

20161
Reset filters

Settings

Citations

Publication Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    Publication
    On the use of embedded devices for heterogeneous cloud infrastructure
    (2016) Kumar, Pardeep; Arzuaga, Emmanuel; College of Engineering; Rivera Vega, Pedro I.; Rodriguez Martinez, Manuel; Department of Electrical and Computer Engineering; Cruz, José
    Regardless of the type of user, be it a computer professional or smart-phone user, working in a small bank or at a high tech IT company, everyone is (knowingly or not) using cloud-computing. In fact, the cloud-computing model provides various advantages over traditional computing in terms of service availability, scalability, processing and administration. Further, the cloud-computing framework relies on the principle of sharing resources and underlying hardware architectures through the virtualization of various components using software abstractions. Increasing popularity of ARM-based boards with multi-core processors, along with commodity hardware components with cost-effective power consumption, yet with smaller and compact design, has exposed a wide range of opportunities to positively impact computing infrastructure design such as redefining building blocks for multiple computing paradigms such as parallel computing, virtual computing, cloud computing, high performance computing, and real-time computing. With widespread availability of such hardware it is possible for the common user to now build a specialized hardware solution to meet requirements on application performance, cost or power consumption. This thesis is undertaken to study and evaluate the use of virtualization to enable the integration of ARM-based embedded boards into cloud computing infrastructure. In particular, this work will focus on the understanding of how effective is virtualization on embedded boards and how this idea could be further nurtured to develop a whole cloud based service that provides for example virtual machines (VMs) on demand. We will discuss various techniques for setting up VMs using open source and publicly available software and tools on ARM-based boards. Our goal is to provide a system that is capable of running common desktop operating systems in current standard architectures, such as MS Windows and Ubuntu Linux on x86 as well as to have the capability of supporting native ARM-based operating systems taking advantage of the full virtualization support using KVM/QEMU on this platform.