Portfolio

About Chathura

Chathura Wanniarachchi graduated from the University of Colombo School of Computing (UCSC), Sri Lanka with a First-Class Computer Science Honours degree. He is passionate about Algorithm Development research as it makes the existing system more efficient and optimized. Apart from that, he has a special interest in drone human interaction-related research which makes drones like cutting-edge technological devices more user-friendly and accessible to the public.

He was selected to the UCSC, completing the GCE Advanced Level examination with an island-wide ranking of 176 from the high school. His interest in research in the computing domain rose during his second year of studies at the university. Initially, it was a search for solutions for the problems found during the year-wide software development group project. Then the Research internship at CoST Research Lab at UCSC focused his interest on algorithm development research. There he developed and Genetics Algorithm-based optimization approach for Audio Fingerprinting algorithms. With that experience, he started the final year of the university with a yearlong thesis-based research project.

During his final year at university, he got a teaching opportunity as an Undergraduate Teaching Assistant. He conducted tutorial and practical sessions Electronics and Physical Computing course module for second-year computer science undergraduates. It helped him to improve his teaching, lecture material making skill and lecturer conducting skills as the student group was well over 200, and they learned utterly new course modules in the curriculum.

My primary research interests focus on developing algorithms for single-threaded, multi-threaded, and distributed computing environments. I am particularly fascinated by integrating biological phenomena into these algorithms to enhance their functionality and efficiency. Additionally, I am deeply interested in edge computing and the design and development of IoT systems. I aim to leverage machine learning to optimize performance on edge devices and process data locally, thereby maintaining data privacy and reducing reliance on cloud-based decision-making.

Apart from all, he is a self-hosting nerd, and he maintains a home lab which provides most of the software solutions he needs to do his day-to-day activities securely with more privacy and full control of his data.

Download Resume

Resume

This section highlights the academic history and professional experiences I gained through different institutions holding different positions and responsibilities.

Education

B.SC.(HONS) IN COMPUTER SCIENCE

January 2019 - May 2023
University of Colombo School of Computing, Sri Lanka
Cumulative GPA - 3.74/4.00
UCSC Director's List Nominee

G.C.E. A/L COMMON STREAM

May 2015 - August 2018
Bandaranayake Central Collage, Veyangoda, Sri Lanka
Island Rank - 176

UP TO G.C.E. O/L

January 2004 - May 2015
Pallewela National Collage, Mirigama, Sri Lanka
Senior Prefect [2012-2015]

Professional Experience

SOFTWARE ENGINEER

May 2023 - Percent
Landon Stock Exchange Group(LSEG)
Distributed system designer and developer for CCP systems

UNDERGRADUATE TEACHING ASSISTANT

November 2022 - March 2023
University of Colombo School of Computing, Sri Lanka
I assisted and conducted lectures and practical sessions for Electronic and Physical Computing course modules of 2nd-year computer science undergraduates.

RESEARCH INTERN

October 2021 - April 2022
University of Colombo School of Computing, Sri Lanka
I studied Content-Based Audio Fingerprinting Algorithms and Genetic Algorithms during the internship program. More specifically, I analyzed and discovered a procedure to optimize hyperparameter values of those algorithms.

Portfolio

The following section will present the research project I engaged with, publications I contributed to, the teaching and mentoring experience I gained via different projects and opportunities, and other miscellaneous projects I working and worked on.

Leader Selection for Single Point of Failure for Single Leader, Leader-Follower Drone Swarm

This was a one-year-long thesis-based research I conducted during my last year of undergraduate under the supervision of Prof. Prasad Wimalarathne and Dr. Kasun Karunanayaka.
The primary goal of this research is to provide a solution to the single point of failure problem attached to the single leader, leader-follower drone swarms. So, I proposed a solution, using the distributed leader election concept with the incorporation of the capability aware measurements.
A step in the proposed algorithm
The proposed algorithm was verified via a series of simulation experiments conducted on the SITL drone simulation environments.
Result of a simulation test case
The implementation of the proposed algorithm using Python3 and DroneKit API can be found in this link.
The following papers are published based on this research,
- Thesis: Click hear
- Research paper published at IEEE IES ONCON23: 10.1109/ONCON60463.2023.10431018
- A review paper published to IEEE ISIE24: 10.1109/ISIE54533.2024.10595819
- Working on the journal paper.

Audio Fingerprinting Algorithm Optimization Using Genetic Algorithms

This research study primarily focused on optimizing content-based audio fingerprinting algorithm hyperparameter to improve the algorithm's overall performance.
Optimized parameter list
I incorporated the CHC genetic algorithm, encoded the hyperparameters into a genetic string, and optimized it with the penalty approach to avoid overfitting cases such as long query length.
Encoded genetic string
The experiments and the research approach helped me understand the conversion process of real-world problems to a genetic algorithm optimizable problem and different methods for audio fingerprinting, including the Shazam algorithm and the state-of-the-art deep learning-based algorithms.
Python3 implementation for the developed optimization algorithm can be found in this link.

HomeLab

The home lab is my simple home server system, which facilitates most of the peripheral computing needs such as self-hosted services and storage services. I have a self-hosted WireGuard VPN on AWS EC2 environment, which helps me to access home services via the public internet without exposing the home network directly to the public.
The base home server system is virtualized using ProxMox, and three virtual machines are sharing the hardware resource. Those three virtual machines have their responsibilities as follows.
1. VM 100 - Responsible for providing storage service using the TrueNAS Core
2. VM 101 - An Ubuntu system work as a docker host to deploy containers used by myself
3. VM 102 - A Debian system works as the internal DNS and the reverse proxy for my services within the network
The system architecture of current home lab setup
This allows me to keep my data with me and manage it more secure manner without handing over personal information to a third party.
Some services available in my home lab

Software Projects

University Student Staff Portal

This is a web application designed by combining MVC and Micro-Services architecture. The server-side language was PHP8, and the web application was developed without any frameworks. So, I develop an MVC framework and a CSS grid system to improve the group development productivity.
I mainly engaged in the design, development, testing and documentation phases of the project. As the team lead, I oversaw the project and delivered it on the given deadline.
Project Repository can be found in this link.
EduPulse

This was another group project I participated in, and we developed it with React(Material UI), Node.js and MongoDB as the database. The service was deployed in Microsoft Azure and MongoDB Atlas.
This was a good opportunity for me to get familiar with cloud environments and software deployment strategies.
Project Repository can be found in this link.
Echole PA

This is a currently ongoing software project developed by myself to accommodate my daily personal needs. This has an in-built scheduler which acts as Linux System Crone and performs scheduled activities. Currently, I am expanding its feature by incorporating matching learning to assist my daily activities.
he system has three major components. They are the backend API, web application using React and a Telegram Bot. All of these services are locally hosted in my home lab, and I regularly made available regular releases to improve the functional and non-functional capabilities.
The code is currently closed source, and repositories are available at the locally hosted Gitea instance.
Public Repositories

Pihole Addlist Aggregator

This repository provides aggregated add lists to the Pihole add blocking utility. It fetches information from https://firebog.net and updates the list daily via scheduled GitHub actions.
Project Repository can be found in this link.

File Sync

This utility can be used to sync files between two systems.
Project Repository can be found in this link.

RDBSync

This is another utility written in Python3, and it facilitates data sync between two relational databases.
Project Repository can be found in this link.

Conference Proceedings

C.Wanniarachchi, P. Wimalaratne, and K. Karunanayaka, "A behavioral model for single leader:Leader-follower drone swarm in leader’s failure," in Proceedings of the 2^nd IEEE Industrial Electronics Society Annual Online Conference (ONCON24), South Carolina, USA, 2023. URL: https://ieeexplore.ieee.org/document/10431018.
C.Wanniarachchi, P. Wimalaratne, and K. Karunanayaka, "Formation Control Algorithms for Drone Swarms and The Single Point of Failure Crisis: A Review" in Proceedings of the 33^rd International Symposium on Industrial Electronics (ISIE24), Ulsan, South Korea, 2024. URL: https://ieeexplore.ieee.org/document/10595819/.
O.Wickramasinghe, N.Rajan, C.Wanniarachchi, K. Karunanayaka and P. Wimalaratne, "Evolution of Formation Control Algorithms for Unmanned Aerial Vehicles" in Proceedings of the 33^rd International Symposium on Industrial Electronics (ISIE24), Ulsan, South Korea, 2024. URL: https://ieeexplore.ieee.org/document/10595718.

Contact

Please reach me via the mentioned email address; it is in the human-only readable format, as I do not wish to talk with machines. Other than that, you can reach me or find out more about me and my work on the mentioned social media and other service links.

LinkedIn | GitHub | ORCiD

hello i'm Chathura

I am a Computer Science Graduate Software Engineer Former Research Intern Former Undergraduate Teaching Assistant

About Chathura

Resume

Education

Professional Experience

Portfolio

Contact