Systems and methods to detect GPS spoofing attacks

Patent metadata

Field	Value
Publication number	US-12445850-B2
Application number	US-202318397610-A
Country	US
Kind code	B2
Filing date	Dec 27, 2023
Priority date	Dec 27, 2023
Publication date	Oct 14, 2025
Grant date	Oct 14, 2025

Abstract

A method for detecting GPS spoofing attacks includes providing a trained deep learning (DL) model based on neural networks, feeding GPS signals into the trained DL model, using asymmetric Shapley values (ASVs) to calculate feature contributions, using the ASVs to assign a non-uniform distribution over an ordering of features, obtaining causal structures among the features, applying the ASVs to causal Shapley additive explanation to obtain Shapley attributions, incorporating the Shapley attributions and the causal structures, and detecting GPS spoofing attacks by running the trained DL model and using the causal structures, non-uniform distribution, feature contributions, and Shapley attributions.

First claim

What is claimed is:

1. A method for detecting a global positioning system (GPS) spoofing attack, comprising:
providing a trained deep learning (DL) model based on a neural network;
feeding a GPS signal into the trained DL model;
using asymmetric Shapley values (ASVs) to calculate a plurality of feature contributions;
using the ASVs to assign a non-uniform distribution over an ordering of a plurality of features;
obtaining a plurality of causal structures among the plurality of features;
applying the ASVs to causal Shapley additive explanation to obtain a Shapley attribution;
incorporating the Shapley attribution and the plurality of causal structures;
using Shapley additive explanation (SHAP) to obtain a reason behind signal classification;
detecting the GPS spoofing attack by running the trained DL model and using the plurality of causal structures, the non-uniform distribution, the plurality of feature contributions, the Shapley attribution, the reason behind the signal classification, and incorporation of the Shapley attribution and the plurality of causal structures;
wherein the neural network includes three hidden layers that are followed by a rectified linear unit (ReLU) activation function or a hyperbolic tangent (Tanh) activation function;
using a Bayesian structural causal model (SCM) to construct a graphical representation of a causal relationship among the plurality of features.

2. The method according to claim 1 , further comprising:
assigning a plurality of values to the plurality of features, respectively.

3. The method according to claim 1 , wherein the trained DL model is stored at a processing chip or a memory connected to the processing chip on an unmanned aerial vehicle (UAV).

4. The method according to claim 1 , further comprising:
using Non-combinatorial Optimization via Trace Exponential and Augmented lagRangian for Structure learning (NOTEARS) to determine the plurality of causal structures.

5. An electronic device for detecting a global positioning system (GPS) spoofing attack, comprising:
one or more processors; and
a memory coupled to the one or more processors and storing computer programs that, when being executed, cause the one or more processors to perform:
providing a trained deep learning (DL) model based on a neural network;
feeding a GPS signal into the trained DL model;
using asymmetric Shapley values (ASVs) to calculate a plurality of feature contributions;
using the ASVs to assign a non-uniform distribution over an ordering of a plurality of features;
obtaining a plurality of causal structures among the plurality of features;
applying the ASVs to causal Shapley additive explanation to obtain a Shapley attribution;
incorporating the Shapley attribution and the plurality of causal structures;
using Shapley additive explanation (SHAP) to obtain a reason behind signal classification;
detecting the GPS spoofing attack by running the trained DL model and using the plurality of causal structures, the non-uniform distribution, the plurality of feature contributions, the Shapley attribution, the reason behind the signal classification, and incorporation of the Shapley attribution and the plurality of causal structures;
wherein the neural network includes three hidden layers that are followed by a rectified linear unit (ReLU) activation function or a hyperbolic tangent (Tanh) activation function;
using a Bayesian structural causal model (SCM) to construct a graphical representation of a causal relationship among the plurality of features.

6. The device according to claim 5 , wherein the one or more processors are further configured to perform:
assigning a plurality of values to the plurality of features, respectively.

7. The device according to claim 5 , wherein the trained DL model and ASV algorithm are stored at the one or more processors or the memory and the device is mounted on an unmanned aerial vehicle (UAV).

8. The device according to claim 5 , wherein the one or more processors are further configured to perform:
using Non-combinatorial Optimization via Trace Exponential and Augmented lagRangian for Structure learning (NOTEARS) to determine the plurality of causal structures.

9. A non-transitory computer readable storage medium, containing computer programs that, when being executed, cause one or more processors of an electronic device to perform:
providing a trained deep learning (DL) model based on a neural network;
feeding a GPS signal into the trained DL model;
using asymmetric Shapley values (ASVs) to calculate a plurality of feature contributions;
using the ASVs to assign a non-uniform distribution over an ordering of a plurality of features;
obtaining a plurality of causal structures among the plurality of features;
applying the ASVs to causal Shapley additive explanation to obtain a Shapley attribution;
incorporating the Shapley attribution and the plurality of causal structures;
using Shapley additive explanation (SHAP) to obtain a reason behind signal classification;
detecting the GPS spoofing attack by running the trained DL model and using the plurality of causal structures, the non-uniform distribution, the plurality of feature contributions, the Shapley attribution, the reason behind the signal classification, and incorporation of the Shapley attribution and the plurality of causal structures;
wherein the neural network includes three hidden layers that are followed by a rectified linear unit (ReLU) activation function or a hyperbolic tangent (Tanh) activation function;
using a Bayesian structural causal model (SCM) to construct a graphical representation of a causal relationship among the plurality of features.

10. The storage medium according to claim 9 , wherein the one or more processors are further configured to perform:
assigning a plurality of values to the plurality of features, respectively.

11. The storage medium according to claim 9 , wherein the one or more processors are further configured to perform:
using Non-combinatorial Optimization via Trace Exponential and Augmented lagRangian for Structure learning (NOTEARS) to determine the plurality of causal structures.

Assignees

• Intelligent Fusion Tech Inc

Inventors

• Tian Xin
• Fan Zhengyang
• Pham Khanh
• Blasch Erik
• Wei Sixiao
• Shen Dan
• Chen Genshe

Classifications

• H04W12/121 Primary

  Wireless intrusion detection systems [WIDS]; Wireless intrusion prevention systems [WIPS] · CPC title