Ken Subratie

Ken is a doctoral candidate in the College of Engineering (Electrical and Computer Engineering) at the University of Florida. He is currently a Research Assistant in the Advanced Computing and Information Systems (ACIS) Laboratory under the guidance of Dr. Renato Figueiredo. His research interests lie in storage systems, virtual networks, and software system design.

I am currently involved in the investigation of virtual networks, distributed storage and resource access control within the context of IoT and edge networks. My research seeks to provide the system software infrastructure necessary to facilitate the rapid and convenient deployment of ad hoc networked groups with enforced access control to shared, distributed resources.
Additionally, I collaborate with the inter-disciplinary research group PRAGMA-GLEON to design and build distributed computing infrastructure and workflows for running intensive lake ecology simulations which better enables research into the dynamics of lake communities. The result of this work is the GRAPLEr project.
I am also recently affiliated with CENTRA (Collaborations to Enable Transnational Cyberinfrastructure Applications) through ACIS Labs. I will investigate an application of my research project through a collaboration with other member researchers, to implement a system for the partnering of scientific data among the Mekong river countries in South East Asia.


IPoP VPN – IPOP (IP-over-P2P) is an open-source user-centric software virtual network allowing end users to define and create their own virtual private networks (VPNs). IPOP virtual networks provide end-to-end IP tunneling over “TinCan” links setup and managed through a control API to create various software-defined VPN overlays.

GRAPLEr brings the power of distributing computing to the fingertips of lake ecology modelers. It is a suite of software tools for connecting to the GRAPLE Web Service (GWS) and running GRAPLE jobs. Experiments consisting of thousands of simulations can be evaluated in a fraction of the time by submitting batches of GLM (General Lake Model) simulations to a distributed computing system, directly from your R/RStudio environment

Intel Labs May – Aug 2016
Research Intern

As a member of the Infrastructure Research Lab I investigated and developed a hypervisor component which used introspection techniques to virtualize performance characteristics of a storage host bus adapter. This component simplifies investigating the impact of storage on contemporary workloads by modeling various storage capabilities within a single device. My contribution was peer-reviewed and accepted for integration into the core code base.

IBM Research May – Aug 2015
Research Intern

As part of the Cloud Storage, Distributed Data Management Team, I was involved in the research of an efficient method to select groups of deduplicated data objects in a clustered environment for distribution across its storage nodes. The result of this work has passed the first stages of review for consideration of patent application with the United States Patent and Trademark Office.