Year: 2025
Type: article
Abstract: The research work offers a unique -created building for reconstructing compressively sensed signals using the Orthogonal Matching Pursuit (OMP) method. This research work investigated the difficulty of computation and data dependencies among the various stages of the OMP algorithm in order to build a structure that gives greater speed while consuming less space. The present paper presents an enhanced OMP technique and its employment using Xilinx Vivado High-Level Fabrication (VHLF). This study employs Gram-Schmidt orthogonalization to enhance signalling residue updating so that sign restoration individual requirements to run the lowest-squares explanation once, subsequent in a substantial decrease in the overall sum of vector processes in an electronic design. Results from simulations demonstrate that our OMP method is similarly precise in signal restoration as the initial OMP technique. Our strategy offers a quick and adaptable deployment for a variety of signal sizes, measuring grid sizes, and sparseness levels. The suggested approach may recuperate a 128-length indicate with evaluation the amount M = 32 and sparseness K = 8 and K = 5 in 21 and 13.2 μs, indicating at least a 20.9% and 21.3% enhancement contrasted against the present HLS-based works. The findings are comparable to those of typical hardware description language (HDL) employments. The findings demonstrate, enhanced OMP method not merely has a faster rebuilding time compared to comparable current HLS-based works, but it can additionally contend with previous attempts developed via the classic FPGA construction path.
DOI: 10.46632/jeae/2/2/20
Language: English
Subfield: Computational Mechanics
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4412506496