GAC-KAN: An Ultra-Lightweight GNSS Interference Classifier for GenAI-Powered Consumer Edge Devices

TL;DR

The paper tackles GNSS interference classification on GenAI-powered edge devices with stringent resource constraints. It introduces GAC-KAN, a physics-guided framework that combines a Multi-Scale Ghost-ACB backbone with a Kolmogorov-Arnold Network head and leverages physics-based generative simulation to address data scarcity. Empirically, it achieves 98.0% accuracy with only 0.13M parameters and low compute (0.19G FLOPs), outpacing state-of-the-art baselines and enabling an always-on GNSS security companion. The approach demonstrates strong robustness across jamming-to-noise scenarios and clear inter-class discriminability, making it practical for secure GenAI-enabled consumer electronics.

Abstract

The integration of Generative AI (GenAI) into Consumer Electronics (CE)--from AI-powered assistants in wearables to generative planning in autonomous Uncrewed Aerial Vehicles (UAVs)--has revolutionized user experiences. However, these GenAI applications impose immense computational burdens on edge hardware, leaving strictly limited resources for fundamental security tasks like Global Navigation Satellite System (GNSS) signal protection. Furthermore, training robust classifiers for such devices is hindered by the scarcity of real-world interference data. To address the dual challenges of data scarcity and the extreme efficiency required by the GenAI era, this paper proposes a novel framework named GAC-KAN. First, we adopt a physics-guided simulation approach to synthesize a large-scale, high-fidelity jamming dataset, mitigating the data bottleneck. Second, to reconcile high accuracy with the stringent resource constraints of GenAI-native chips, we design a Multi-Scale Ghost-ACB-Coordinate (MS-GAC) backbone. This backbone combines Asymmetric Convolution Blocks (ACB) and Ghost modules to extract rich spectral-temporal features with minimal redundancy. Replacing the traditional Multi-Layer Perceptron (MLP) decision head, we introduce a Kolmogorov-Arnold Network (KAN), which employs learnable spline activation functions to achieve superior non-linear mapping capabilities with significantly fewer parameters. Experimental results demonstrate that GAC-KAN achieves an overall accuracy of 98.0\%, outperforming state-of-the-art baselines. Significantly, the model contains only 0.13 million parameter--approximately 660 times fewer than Vision Transformer (ViT) baselines. This extreme lightweight characteristic makes GAC-KAN an ideal "always-on" security companion, ensuring GNSS reliability without contending for the computational resources required by primary GenAI tasks.

Figures (5)

• Figure 1: Time-Frequency representations (spectrograms) of the generated jamming primitives considered in the system model. The horizontal axis represents time ($\mu s$) and the vertical axis represents frequency (MHz).
• Figure 2: The overall architecture of the proposed GAC-KAN (left) and the detailed structure of the Ghost-ACB-CA Unit (right). The network processes STFT spectrograms through stacked MS-GAC blocks and employs a KAN head for classification.
• Figure 3: Structure of the Multi-Scale Ghost-ACB-Coordinate (MS-GAC) Block utilizing parallel branches with varying receptive fields.
• Figure 4: Classification accuracy versus Jamming-to-Noise Ratio (JNR). The proposed GAC-KAN demonstrates superior robustness in low JNR regimes (-25 dB to -15 dB).
• Figure 5: Confusion Matrix of the GAC-KAN model on the test dataset. The vertical axis represents the true labels, and the horizontal axis represents the predicted labels.