Floating Point Implementation of the Improved QRD and OMP for Compressive Sensing Signal Reconstruction

Alahari Radhika; Kodati Satya Prasad; Kalitkar Kishan Rao

首页> 外文期刊>Sensing and imaging >Floating Point Implementation of the Improved QRD and OMP for Compressive Sensing Signal Reconstruction

【24h】

Floating Point Implementation of the Improved QRD and OMP for Compressive Sensing Signal Reconstruction

机译：Floating Point Implementation of the Improved QRD and OMP for Compressive Sensing Signal Reconstruction

获取原文

获取原文并翻译 | 示例

掌桥外文数据库（机构版） >>

开具论文收录证明 >>

文献代查 >>

页面导航

摘要
著录项
相关主题

摘要

In this paper, the Floating-Point Core Architecture based QR decomposition is proposed for solving least square problems in the Orthogonal Matching Pursuit algorithm (OMP-FPCA-QRD). To improve the computational performance of Orthogonal Matching Pursuit (OMP), it is necessary to modify the Orthogonal Matching Pursuit algorithm for analysing a wide range of signals in field programmable gate array (FPGA). As a result, it highly benefits from the available resources and acquires a scalable computational complexity. Since the solution of least square problem involves some iterative parts, like square root and division units, the processing time of the proposed QR Decomposition (QRD) approach is decreased by increasing parallelism using processing element driven systolic array implementation across all data-dependent operations. The hardware implementation on the ALTERA field programmable gate array shows optimal performance depends on hardware complexity and frequency of operation with the improved computational accuracy over existing QR Decomposition implementations. Moreover, the implementation of Orthogonal Matching Pursuit algorithm for signal reconstruction is also proposed to validate the performance metrics of floating point unit (FPU). The experimental results show that the optimization of floating point unit offers significant resource optimization in QR decomposition, and also better performance of high peak signal-to-noise ratio of 32.99 dB, which outperforms all other fixed point Orthogonal Matching Pursuit systems.

著录项

来源
《Sensing and imaging》 |2022年第1期|共17页
作者
Alahari Radhika; Kodati Satya Prasad; Kalitkar Kishan Rao;
展开▼
作者单位

JNT Univ;

UCEK;

NITW;

展开▼
收录信息
原文格式 PDF
正文语种英语
中图分类仪器、仪表;
关键词
Compressive sensing; QRD implementation; Hardware fusion; Orthogonal Matching Pursuit; Field programmable gate array; ORTHOGONAL MATCHING PURSUIT; FPGA IMPLEMENTATION; RECOVERY;

Floating Point Implementation of the Improved QRD and OMP for Compressive Sensing Signal Reconstruction

摘要

著录项

相关主题

期刊订阅