GNSS Satellite Selection-based on Per-satellite Parameters Using Deep Learning

Abstract

In recent years, there has been a drastic increase in navigation satellites. Due to the computational limitations of a global navigation satellite system (GNSS) receiver, such as limited tracking channels, bandwidth, and high-power consumption, selecting an optimal subset of satellites with a better geometric dilution of precision (GDOP) has become increasingly difficult. Moreover, the computation load on the receiver increases with multi-channel observation. However, a reduced subset of used satellites will significantly increase the receiver computation performance without compromising the positioning accuracy. This paper presents an end-to-end neural network (NN) architecture based on PointNet and VoxelNet to provide an optimal satellite subset from a multi-constellation with a minimum GDOP. The architecture is based on per-satellite parameters obtained from the global positioning system (GPS) and navigation with the Indian Constellation (NavIC) satellites. Experimental results indicate that the input channel with satellite positions, Doppler, acquisition peak metric, elevation, and azimuth angles provides the best outcome in terms of performance. This method can reduce the computation burden and time by 40–50% when used in the receiver.

KEYWORDS:

• Deep learning
• GPS
• NavIC
• Satellite selection
• Software receiver
• Universal software radio peripheral (USRP)

ACKNOWLEDGEMENTS

The universal software radio peripheral (USRP)-2932 devices used in this work were purchased under the research contingency fund from Indian Space Research Organization (ISRO) [grant number ISRO/RES/3/720/2016-17]. The software receiver used for preprocessing the recorded NavIC and GPS signals was developed under the ISRO RESPOND project, sanctioned by Space Application Center (SAC), Ahmedabad. We would like to express our gratitude to the Space Applications Center (SAC) – Ahmedabad, ISRO, for all the assistance provided to ensure the successful completion of the project.

Additional information

Funding

This work was supported by the Indian Space Research Organization (ISRO) project under grant number ISRO/RES/3/720/2016-17.

Notes on contributors

Prateek Singh

Prateek Singh is a PhD researcher in the Department of Electrical and Electronics Engineering, Birla Institute of Technology and Science-Pilani, K K Birla Goa Campus, India. He received his Master's of Technology degree in electronics and communication engineering from Guru Gobind Singh Indraprastha University, Delhi, India, in 2017. He worked as a junior research fellow in the Indian Space Research Organization (ISRO) funded project from August (2017–2019). His current research focus is on signal processing of software-defined GNSS receivers. Corresponding author. Email: [email protected]

Janamejay Joshi

Janamejay Joshi graduated from Birla Institute of Technology and Science, Pilani – Goa Campus, India in 2021, with a dual degree in MSc (Mathematics) and BE (Computer Science). He is working on interdisciplinary problem statements that leverage the power of data. With a data science based profile, his work so far spans a variety of domains – healthcare, autonomous driving systems, engineering consulting, water distribution networks, intelligent hiring solutions and remote sensing. His research interests include media analytics and sports data science. Email: [email protected]

Abhijit Dey

Abhijit Dey received the BE degree from the Department of Electronics and Communication Engineering, North Maharashtra University, India in 2011. In 2014, he received the master's degree in radio frequency and microwave engineering. He is currently a PhD student in the Department of Electrical and Electronics Engineering, Birla Institute of Technology and Science, Pilani, India, before he worked as a junior research fellow in the same institute. He was a visiting PhD researcher with the Department of Aeronautical and Aviation Engineering, The Hong Kong Polytechnic University, Hong Kong, China in 2021. His research interests include signal processing in software-defined GNSS receivers, ionospheric scintillation detection and mitigation techniques in GNSS systems. Email: [email protected]

Nitin Sharma

Nitin Sharma received his PhD from Birla Institute of Technology and Science Pilani, India in 2014 and his research dissertation is in wireless communication. He is also serving as an associate professor in the Department of EEE at BITS Pilani K.K. Birla Goa Campus. His current research interests include optimization in wireless communication, resource allocation and management for wireless and wired communications, cognitive radio, green communication technologies, M2M/IoT communications, cloud communication and evolutionary computations. Dr Nitin is also working in the area of GNSS signal processing and is currently working as the principal investigator of a funded project from ISRO in this area. Dr Nitin has more than 15 publications in peer-reviewed journals in the area of optimization in wireless data transmission. He also serves as an associate editor for three international journals and has been an active reviewer for many journals that include IEEE TEC, IEEE Systems, Elsevier's Computers and Electrical Engineering, Information Sciences, Computers and Electrical Engineering, Springer's Wireless Personal Communications and Frontiers of Computer Sciences. He is also serving as a TPC member for prestigious IEEE conferences such as ICC and GLOBECOM. Email: [email protected]