AI-Driven "Security-by-Design" Router for Anomalous Traffic Detection

We are developing a deep learning-based baseline for a next-generation home router that proactively detects network anomalies using behavioral analysis. Unlike signature-based systems, our model utilizes CNNs (Convolutional Neural Networks) to process raw packet data transformed into fixed-size images, enabling the detection of both known and zero-day attacks.

• Hardware-Accelerated: Optimized for edge deployment on NVIDIA Jetson AGX Orin, allowing for real-time inference and on-device retraining.