Investigating SAMV Regarding its Suitability For FPGAs

Giahi F, Rachuj S, Fey D (2022)

Publication Type: Conference contribution

Publication year: 2022

Original Authors: Farehe Giahi, Sebastian Rachuj, Dietmar Fey

Series: Institute of Electrical and Electronics Engineers and Institution of Engineering and Technology (IEEE Xplore)

City/Town: Belfast, United Kingdom

Conference Proceedings Title: 2022 IEEE 35th International System-on-Chip Conference (SOCC)

Event location: Belfast GB

DOI: 10.1109/SOCC56010.2022.9908124

Abstract

SAMV is an iterative algorithm for determining the direction of arrival (DOA) usable for different sensor technologies, like Radar. It offers more benefits than conventional algorithms used for the same goal like MUSIC and ESPRIT. However, it is more complex and requires more computational performance. Since DOA algorithms are usually employed in embedded systems, dedicated accelerator cores are required for using SAMV in these applications. As there are no known investigations available that look into implementing SAMV as discrete hardware, this paper aims to provide an analysis of possible implementations. For this, different designs for an FPGA are proposed and studied. FPGAs are becoming more widespread in embedded systems and are often already integrated in common MPSoCs. That makes them a decent platform for very specialized accelerator cores. It was feasible to show an improvement in speed in range of three to seven times in comparison to a CPU implementation.

Authors with CRIS profile

Farehe Giahi Lehrstuhl für Informatik 3 (Rechnerarchitektur) Sebastian Rachuj Lehrstuhl für Informatik 3 (Rechnerarchitektur) Dietmar Fey Lehrstuhl für Informatik 3 (Rechnerarchitektur)

How to cite

APA:

Giahi, F., Rachuj, S., & Fey, D. (2022). Investigating SAMV Regarding its Suitability For FPGAs. In 2022 IEEE 35th International System-on-Chip Conference (SOCC). Belfast, GB: Belfast, United Kingdom.

MLA:

Giahi, Farehe, Sebastian Rachuj, and Dietmar Fey. "Investigating SAMV Regarding its Suitability For FPGAs." Proceedings of the 2022 IEEE 35th International System-on-Chip Conference (SOCC), Belfast Belfast, United Kingdom, 2022.

BibTeX: Download