Date of Award
10-5-2017
Publication Type
Master Thesis
Degree Name
M.A.Sc.
Department
Electrical and Computer Engineering
Keywords
automotive, FPGA, High Level Synthesis, RADAR, Vivado HLS, Xilinx
Supervisor
Khalid, Mohammed
Rights
info:eu-repo/semantics/openAccess
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Abstract
High Level Synthesis (HLS) is a technology used to design and develop hardware (HW) using high-level languages such as C/C++. An HLS model of an automotive RADAR signal processing algorithm has been developed for the purpose of comparison between the HLS model and the existing HDL model. Register Transfer Level (RTL) programming is a technology used to design and develop hardware at the register transfer level (or low level) using Hardware description languages such as Verilog and VHDL. FPGA development usually requires the knowledge of RTL technologies. HLS gives software (SW) developers the ability to design and implement their designs on an FPGA without requiring the knowledge of RTL technologies and HDL. Even though HLS is currently gaining popularity, the applications used to evaluate HLS tend to remain small. We synthesize an automotive RADAR signal processing system using HLS-based design methodology, which has mid to high complexity, and compare our synthesis results to that of the RTL-based design. We used many techniques used to make the high-level program model ready for synthesis while optimizing for both speed and resource usage using Xilinx Vivado HLS Computer-Aided Design (CAD) tool. We achieved a speed up of 2X compared to the RTL-based design while reducing the design time from approximately 16 weeks to 6 weeks. The FPGA resource utilization increased but it was still under 5% of the total resources available on the FPGA.
Recommended Citation
Luthra, Siddhant, "High Level Synthesis and Evaluation of an Automotive RADAR Signal Processing algorithm for FPGAs" (2017). Electronic Theses and Dissertations. 7274.
https://scholar.uwindsor.ca/etd/7274