FPGA Programming Challenges When Estimating Power Spectral Density and Autocorrelation in Coherent Doppler Lidar...

Area of Science:

• Engineering
• Atmospheric Science
• Signal Processing

Background:

• Coherent Doppler Lidar (CDL) systems are crucial for remote wind sensing.
• Efficient hardware implementation of signal processing algorithms is essential for real-time CDL applications.

Purpose of the Study:

• To present FPGA logic designs for two distinct signal pre-processing algorithms for CDL wind sensing.
• To address and overcome hardware implementation challenges associated with these algorithms.
• To demonstrate the efficacy of the implemented algorithms through atmospheric wind measurements.

Main Methods:

• Developed two FPGA hardware programming algorithms: a segmented spectral analysis and an autocorrelation-based approach.
• Utilized Xilinx System Generator for logic design and VLSI image creation.
• Implemented algorithms on an FPGA, addressing challenges like bit growth, memory constraints, and signal drive issues.

Main Results:

• Successfully designed and implemented two distinct FPGA algorithms for CDL signal processing.
• Devised solutions for critical hardware implementation challenges including bit growth, memory limitations, and signal drive.
• Obtained atmospheric wind measurements validating the performance of the developed algorithms.

Conclusions:

• FPGA implementation of CDL signal processing algorithms is feasible with appropriate design considerations.
• The presented algorithms and their hardware solutions enable effective wind sensing using Coherent Doppler Lidar.
• The study provides valuable insights into optimizing hardware acceleration for remote sensing applications.