Honors Program Theses

Award/Availability

Open Access Honors Program Thesis

First Advisor

Andrew Berns

Abstract

Wireless sensor networks (WSNs) are widely used in fields such as environmental monitoring, smart agriculture, disaster response, and other applications where low-power devices must communicate reliably over time. Because these networks operate under limited energy and communication constraints, routing and broadcast strategies strongly influence performance, scalability, and network lifetime. This thesis examines how simulation and machine learning can be used to study interference-aware broadcasting in WSNs. The project first evaluates the usefulness of existing network simulation tools, including ns-3, and then describes the development of a custom Python simulator designed for rapid experimentation with broadcast behavior, mobility, and signal-to-interference-plus-noise ratio (SINR)-based reception. The resulting simulator of this project supports fixed and probabilistic transmission, multiple mobility models, and repeated experimental runs across increasing network sizes. A custom Gymnasium environment implemented with Stable-Baselines3 was then used to explore reinforcement learning for parameter tuning, particularly the selection of transmission probability under the SINR model that improves performance compared with fixed-probability baselines. Results show that reinforcement learning can reduce total transmission cost in several larger-network cases, although convergence speed and packet delivery rate remain sensitive to node density, mobility pattern, and interference. Overall, the project demonstrates that a custom simulation workflow can provide a useful foundation for evaluating WSN broadcast strategies and for future work on adaptive learning-based routing.

Year of Submission

2026

Department

Department of Computer Science

University Honors Designation

A thesis submitted in partial fulfillment of the requirements for the designation University Honors

Date Original

2026

Object Description

1 PDF file (23 pages)

Language

en

File Format

application/pdf

Download

Comments?

Share

COinS