Our recent publication :).
In grappling with the intricacies of hardware, a key hurdle emerged concerning complex numbers – especially in metaheuristics optimization algorithms. Our binary approach, marked by precision and adaptability, elegantly addressed this challenge, offering a streamlined solution for efficient FPGA implementation.
In this research we delve into the world of cutting-edge metaheuristic algorithms, gaining popularity for their prowess in solving complex optimization problems.
Curious about their potential for FPGA implementation, we embarked on a journey to identify and adapt suitable metaheuristics. Our spotlight fell on the Simulated Kalman Filter (SKF) – a low-complexity, minimal-step algorithm.
The adventure continued as we transformed the original SKF into the Discrete SKF, rounding off floating-point values and fixing the Kalman gain at 0.5. But that wasn’t the end! We harnessed the power of behavioral modeling to birth the Binary SKF, fine-tuned for FPGA implementation.
The design strategy emphasized modularity, breaking down the metaheuristic into distinct modules for efficient management. Plus, we optimized with a Parallel-In-Parallel-Out configuration for ports.
Simultaneously, we simulated the Discrete SKF on MATLAB and brought the Binary SKF to life on FPGA. The results? Mind-blowing! The Binary SKF achieved speeds up to 69 times faster than the Discrete SKF simulation.
Join us in celebrating this breakthrough 🙂 , where FPGA meets metaheuristic excellence!
#UMPSTEMLab #FPGA #MetaheuristicAlgorithms