2022


Federated Learning For WIFI Fingerprinting

Technologies:

Research project involving applying state-of-the-art ML learning methods to the field of wireless communications. Research was conducted at the University of Toronto WirLab lead by Prof. Shahrokh Valaee and under the guidance of a PhD student.

First author on a paper published at the International Conference on Communcations 2022 in Seoul, South Korea.

Federate Learning For WIFI Fingerprinting Paper Link

2021


Assesing The Scalability Of Semantic-Aware Resource Characterization and Prediction (SAWCAP) Algorithm

Technologies:

Final year project completed in a team of 4 undergraduate students under the supervision of Professor Cristiana Amza.

Project involved assesing the scalability of SAWCAP algorithm on a large-scale Hadoop cluster (> 100 Virtual Machines)

SAWCAP Paper

Github Repository

Physics Simulator

Technologies:

A newtonian physics simulator built in Rust. The main goal of the project was to learn the Rust programming language.

Particles are modelled as circles in 2D space. Collision detection is used to simulate elastic collisions.

Github Repository

Executing Assembly With Deep Learning

Technologies:

A sequence-to-sequence LSTM model to run a sequence of assembly instructions and output the final register states. More can be found in the project’s Github repository.

Github Repository

Course Project: Distributed Key-Value Store

Technologies:

Project done as part of Winter 2021 ECE419 Distributed Systems course.

A distributed key-value store implemented from scratch in Java. The database uses consistent hashing to distribute keys among the nodes in the cluster. Replication is used to achieve fault-tolerance.

Course Project: Waste Classification

Technologies:

Project done as part of Winter 2021 APS360 Artificial Intelligence Fundamentals course.

A CNN model to categorize waste as organic or recyclable.

Github Repository