Graduate Project
 

Deep Learning Based Automatic Modulation Classication for Wideband Access Using Cyclostationarity Analysis

Public Deposited

Downloadable Content

Download PDF
https://ir.library.oregonstate.edu/concern/graduate_projects/8336h822w

Descriptions

Attribute NameValues
Creator
Abstract
  • RF-based signal identification and classification has received growing attention during recent years due to its potential use in many application domains. Of particular interest is Automatic Modulation Classification (AMC), which has been useful in addressing various spectrum related challenges such as signal jamming, policy enforcement, and spectrum sharing. Adopting AMC to wideband spectrum access presents, however, several practical challenges, mainly due to the high sampling rate, computational cost, and memory requirements, which make it unsuitable for real-time scenarios. To address these challenges, we combine the merits of cyclostationarity features and Convolutional Neural Networks (CNN) to propose an efficient wideband AMC technique. Our technique leverages Spectral Correlation Function (SCF) analysis to robustly identify the modulation type of the occupied signal in each of the channels of the wideband spectrum. Our technique does not require prior knowledge of signal parameters such as carrier frequency, symbol rate, and phase offset. We show that our technique outperforms IQ based classifier in terms of accuracy especially under a severe fading environment and requires less training time to converge. Finally, to reduce the cost furthermore, we establish a compressed learning scheme using a few measurements obtained by linear projections of the input features.
Contributor
License
Resource Type
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Rights Statement
Publisher
Peer Reviewed
Language
Embargo reason
  • Ongoing Research
Embargo date range
  • 2020-09-11 to 2021-10-11

Relationships

Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
file details SeifeddineMEJRI-2020.pdf 2020-09-10 Public Download