DeePoly combines rough neural-network features with local polynomial bases and a linear solve to reach high-order accuracy on function approximation and PDEs.
A GPS spoofing detection and classification correlator-based technique using the LASSO
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
This work proposes a global navigation satellite system (GNSS) spoofing detection and classification technique for single antenna receivers. We formulate an optimization problem at the baseband correlator domain by using the Least Absolute Shrinkage and Selection Operator (LASSO). We model correlator tap outputs of the received signal to form a dictionary of triangle-shaped functions and leverage sparse signal processing to choose a decomposition of shifted matching triangles from said dictionary. The optimal solution of this minimization problem discriminates the presence of a potential spoofing attack peak by observing a decomposition of two different code-phase values (authentic and spoofed) in a sparse vector output. We use a threshold to mitigate false alarms. Furthermore, we present a variation of the minimization problem by enhancing the dictionary to a higher-resolution of shifted triangles. The proposed technique can be implemented as an advanced fine-acquisition monitoring tool to aid in the tracking loops for spoofing mitigation. In our experiments, we are able to distinguish authentic and spoofer peaks from synthetic data simulations and from a real dataset, namely, the Texas Spoofing Test Battery (TEXBAT). The proposed method achieves 0.3% detection error rate (DER) for a spoofer attack in nominal signal-to-noise ratio (SNR) conditions for an authentic-over-spoofer power of 3 dB.
citation-role summary
citation-polarity summary
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
DeePoly: A High-Order Accuracy Scientific Machine Learning Framework for Function Approximation and Solving PDEs
DeePoly combines rough neural-network features with local polynomial bases and a linear solve to reach high-order accuracy on function approximation and PDEs.