Parallel implementation by the FPGA of phase diversity based on an improved particle swarm optimization algorithm.

Area of Science:

• Optics and Photonics
• Computer Engineering
• Aerospace Engineering

Background:

• Phase diversity (PD) algorithms are crucial for wavefront sensing but computationally intensive.
• Population optimization methods, like particle swarm optimization (PSO), are often used but suffer from local extrema and long computation times.
• These limitations make traditional PD algorithms unsuitable for resource-constrained applications like space optical systems.

Purpose of the Study:

• To adapt and optimize the phase diversity algorithm for parallel computing architectures.
• To improve the computational speed and performance of PD algorithms for wavefront sensing.
• To enable the application of PD algorithms in power-limited environments such as space optics.

Main Methods:

• Analysis and improvement of the phase diversity algorithm and its particle swarm optimization (PSO) solver.
• Development of a parallel algorithm architecture suitable for field-programmable gate arrays (FPGAs).
• Implementation and board-level verification of the optimized algorithm on an FPGA.

Main Results:

• The developed FPGA-based solution significantly enhances the computational speed of the phase diversity algorithm.
• The optimized algorithm demonstrates improved performance, overcoming limitations of traditional population optimization methods.
• Successful board-level verification confirms the algorithm's viability for real-world applications.

Conclusions:

• The FPGA implementation provides an efficient and high-performance solution for phase diversity wavefront sensing.
• This work makes advanced wavefront sensing feasible for space optical systems with limited computing power and energy.
• The parallel processing approach on FPGAs offers a robust method for computationally demanding optical applications.