Weak signal detection based on deep learning

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Weak signal detection of radio communication signals in complex background noise is an essential part of modern signal processing science. Despite wide application of classical process in various signal detection tasks, the exclusive filter in terms of background noise of radio channel impedes the deployment on modern complex electromagnetism environment. This study introduces a new method of radio signal detection via convolutional neural network (CNN) and bounding box regression. This approach has improved the recent performance of computer vision for object detection. Numerous experiments have shown that Faster R-CNN can accurately detect signal portion in noise, while achieving high-level contextual understanding with millisecond latency compared to traditional schemes.

Original languageEnglish
Title of host publicationICMSSP 2019 - 2019 4th International Conference on Multimedia Systems and Signal Processing
PublisherAssociation for Computing Machinery
Pages114-118
Number of pages5
ISBN (Electronic)9781450371711
DOIs
StatePublished - 10 May 2019
Event4th International Conference on Multimedia Systems and Signal Processing, ICMSSP 2019 - Guangzhou, China
Duration: 10 May 201912 May 2019

Publication series

NameACM International Conference Proceeding Series

Conference

Conference4th International Conference on Multimedia Systems and Signal Processing, ICMSSP 2019
Country/TerritoryChina
CityGuangzhou
Period10/05/1912/05/19

Keywords

  • Bounding box regression
  • Convolutional neural network
  • Signal detection

Fingerprint

Dive into the research topics of 'Weak signal detection based on deep learning'. Together they form a unique fingerprint.

Cite this