Implementation and testing of a GNSS system consisting of a RF front-end and a software GNSS receiver

Table of Contents

1. Introduction and targets

2 Theory

2.1 Software-Defined Radio - SDR

2.2 Software GNSS Receiver - SGR

2.2.1 Spectral analysis

2.2.2 Acquisition of GPS C/A signals

2.2.3 Tracking of GPS C/A signals

2.2.4 Navigation message and PVT calculations

3 Methods and results

3.1 Analysis and documentation of the code

3.1.1 M2HTML and GraphViz for Borre’s Matlab code

3.1.2 Adaptions of “initSettings.m” before the start of programs

3.1.3 Call sequence in the Borre code

3.1.4 NSL Stereo hardware

3.1.5 NSL Stereo C-programs

3.1.6 NSL Stereo Matlab programs

3.2 Solutions for missing functionality, quality, and performance

3.2.1 Determination of receiver velocities

3.2.2 Signal quality C/No and stopping tracking of bad channels

3.2.3 Handling of records of more than 37 seconds

3.2.4 Statistics and quality of navigation results

3.2.5 Performance improvements

3.3 New signals (non GPS L1)

4 Evaluation & Outlook

4.1 Functionality and speed of new Matlab Code

4.2 Interpretation of NSL Stereo Measurement of non GPS L1 signals

4.3 Proposals for future work

Research Objectives and Focus

This work aims to integrate the Nottingham Scientific Limited (NSL) "Stereo" Radio Frequency Frontend (RF-FE) with the open-source "SoftGPS" Matlab software developed by Borre (2007). The primary objective is to enable the acquisition and tracking of various GNSS signals, perform signal quality analysis, calculate position, velocity, and time (PVT), and improve the code's modularity and processing speed, particularly through parallelization and continuous processing simulation.

• Integration of NSL Stereo hardware with existing Matlab-based GNSS receiver software.
• Documentation and modular optimization of the Matlab signal processing pipeline.
• Implementation of performance enhancements including parallel computing and continuous processing simulations.
• Evaluation of signal processing for non-GPS L1 signals using the NSL Stereo front-end.
• Statistical assessment of navigation results compared to a reference measurement.

Excerpt from the Book

Summary of Chapters

1. Introduction and targets: This chapter outlines the motivation for replacing dedicated hardware with software-defined radio solutions and defines the primary goals of integrating the NSL Stereo RF-FE with the SoftGPS Matlab codebase.

2 Theory: Provides the theoretical foundations of Software-Defined Radio (SDR) and Software GNSS Receivers (SGR), covering spectral analysis, signal acquisition, tracking techniques, and navigation data decoding.

3 Methods and results: Details the documentation, code analysis, and technical solutions implemented to enhance functionality, modularity, and processing performance, including parallel processing and continuous data handling.

4 Evaluation & Outlook: Summarizes the achievements regarding functionality and performance improvements while discussing limitations in hardware-software integration and providing recommendations for future research and alternative hardware.

Keywords

Software-Defined Radio, GNSS, NSL Stereo, Matlab, Signal Acquisition, Tracking, Parallel Processing, C/A code, Navigation Message, Spectral Analysis, Performance Optimization, Geodesy, RF-FE, Software-Defined GNSS Receiver.

Frequently Asked Questions

What is the primary focus of this research?

The thesis focuses on adapting and extending the open-source "SoftGPS" project by Borre (2007) to work with the NSL "Stereo" hardware, aiming to modernize its processing capabilities and documentation.

What are the main thematic areas covered?

The work covers hardware configuration for NSL Stereo, Matlab code documentation using M2HTML and GraphViz, signal tracking optimization, and statistical analysis of GNSS positioning results.

What is the primary research goal?

The main goal is to successfully integrate the NSL Stereo front-end with Matlab code to perform robust GNSS signal acquisition and tracking, while improving code modularity and execution speed.

Which scientific methods were employed?

The author utilized software engineering practices for code documentation and modularization, along with digital signal processing (DSP) techniques like PLL/DLL lock loops for tracking and least squares methods for position calculation.

What does the main body address?

The main body documents the analysis of the existing code, the adaptations required for the new hardware, solutions for performance bottlenecks, and the evaluation of navigation data quality.

How is the work characterized in terms of keywords?

It is characterized by terms such as Software-Defined Radio, GNSS, Matlab, signal tracking, parallel processing, and performance optimization.

Why was Git chosen for version control?

Git was selected as it is more modern, widely distributed, and integrated into Matlab, allowing for comprehensive change history tracking compared to the older CVS system.

How did parallel processing impact the system performance?

By implementing "parfor" loops and delegating work to multiple background processes, the CPU utilization increased effectively, significantly reducing the tracking time for a 37-second cycle.

What are the limitations regarding non-GPS L1 signals?

The acquisition of signals like GLONASS or Galileo using the NSL Stereo hardware and software showed deficiencies, likely due to hardware or configuration issues rather than the Matlab code itself.