GPS-Denied Navigation using a Deep Learning Convolutional Neural Network (DL-CNN)

In contested environments, navigation will be compromised due to the ease of jamming the low-power, distributed GPS satellite signals. The Air Force has recognized the vulnerability of military hardware to GPS jamming and spoofing. Current alt-NAV solutions have significant drawbacks. Vision-NAV requires good weather and terrestrial operation. Star tracking requires good weather and only operates at night. Exploiting RF signals of opportunity (such as FM radio of digital TV signals) for navigation is low accuracy and requires a dedicated datalink. Alternate satellite constellations can also be easily jammed (e.g., NTS-3). Our aircrew require precise positioning, navigation, and timing information resilient to any denial method to ensure success against emerging peer competition! ISL’s innovative solution detailed below directly supports the National Defense Strategy by providing “…Joint Lethality in Contested Environments.” In addition, it addresses a Strategic Challenge of the Air Force Science & Technology Strategy to “Provide self-contained, precise positioning, navigation, and timing resilient to any jamming.” We have developed an alt-NAV approach that is all weather and can operate over water. The overall operation and current performance have been validated in simulation. The proposed Phase I program will improve the geolocation accuracy of the alt-NAV approach.