Joshua Agarth

Mentor: Dr. John Shea
College of Engineering
"I got involved with research to gain hands-on experience and to broaden my understanding of different fields in Electrical Engineering. My current research environment allows for collaboration with graduate students and professors. The underlying goal of research is to ask questions nobody has asked before and to search for answers. The search for knowledge has always been a motivating factor in my research."


Electrical Engineering



Research Interests

  • Signals and Systems
  • Wireless Networking
  • Robotics

Academic Awards

  • Charles Poekel Engineering Scholarship
  • Florida Bright Futures
  • University Scholars Program 2016


  • Cru at UF
  • IEEE


  • Greenhouse Church
  • Cru at UF

Hobbies and Interests

  • Playing Guitar
  • Coding
  • Electronic Systems

Research Description

GPU Processing for Cognitive Radio using Machine Learning
The goal of my research is to implement methods for error-correction coding into a graphics-processing unit (GPU) using NVidia’s parallel computing platform, CUDA. GPU’s are multi-core systems that allow effective manipulation of large data blocks and simultaneous computations. I believe implementing processes such as convolutional coding to GPU's can greatly enhance the functionality and speed in a network of autonomous radios.

Enhancing technology calls for communication systems to be able to adapt on multiple dimensions. I would like to develop a system dependent on machine-learning so the radio can understand its conditions and make decisions to be a better peer in a cooperative radio environment. The machine-learning algorithm will be able to change and adapt on various dimensions including time, space, frequency, and code.