Real-Time Phishing Detection Using Microcontroller and Machine Learning

Abstract

Phishing attacks pose a significant threat to online security by tricking users into revealing sensitive information through deceptive website. This project proposes a real-time phishing detection system that integrates machine learning with a microcontroller-based platform to provide a fast, low cost and automated solution. The system uses an ESP32 microcontroller to capture URLs in real-time and transmit them to a machine learning model hosted on a cloud-based Flask server. The model analyzes each URL to classify it as either phishing or legitimate based on learned patterns and features. Experiment results demonstrate the system’s effectiveness in detecting phishing websites with high accuracy and responsiveness, using visual alerts via an OLED display and physical warning through LEDs and buzzers. The proposed approach offers a portable and scalable cybersecurity solution especially useful for environments with limited computing resources. Future enhancements may include the integration of deep learning models and mobile interfaces for wider usability.