CycloDSP: A cyclostationary signal analysis tool for GNU Radio

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Darsena, Donatella, Gelli, Giacinto, Iudice, Ivan
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909211247509504
author Darsena, Donatella
Gelli, Giacinto
Iudice, Ivan
author_facet Darsena, Donatella
Gelli, Giacinto
Iudice, Ivan
contents In this paper, we present a first attempt to incorporate in the GNU Radio ecosystem a tool called CycloDSP, devoted to the analysis of complex-valued cyclostationary signals. Such signals are ubiquitous in communication and signal processing, exhibiting periodic or almost periodic statistics that are characterized by a countable set of cycle frequencies, which are related to the main signal periodicities. Common cycle frequencies for modulated signals are multiple of the baudrate and/or combination of the carrier frequency and baudrate. Conventional estimation strategies for cyclostationary signal analysis typically exhibit a high computational burden. Many approaches aimed at reducing complexity exploit fast Fourier transform (FFT) algorithms, which are very efficient for batch data but are not suited for continuously streaming data, typically encountered in software-defined radio (SDR) applications. The aim of this paper is to develop an out-of-tree (OOT) GNU Radio module containing a set of building block functions, aimed at estimating functions, such as the cyclic correlation functions, typically employed for cyclostationary signal analysis. The proposed implementation must be designed so as to ensure high efficiency in processing continuous data streams of complex samples, both in terms of computational load and data storage, in order to be executed on general purpose computers. We tested our implementation by estimating in real-time the second-order cyclic statistics of a Gaussian minimum-shift keying (GMSK) modulated signal, which belongs to a class of signals commonly employed in aeronautical telemetry datalinks.
format Preprint
id arxiv_https___arxiv_org_abs_2405_16911
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle CycloDSP: A cyclostationary signal analysis tool for GNU Radio
Darsena, Donatella
Gelli, Giacinto
Iudice, Ivan
Signal Processing
In this paper, we present a first attempt to incorporate in the GNU Radio ecosystem a tool called CycloDSP, devoted to the analysis of complex-valued cyclostationary signals. Such signals are ubiquitous in communication and signal processing, exhibiting periodic or almost periodic statistics that are characterized by a countable set of cycle frequencies, which are related to the main signal periodicities. Common cycle frequencies for modulated signals are multiple of the baudrate and/or combination of the carrier frequency and baudrate. Conventional estimation strategies for cyclostationary signal analysis typically exhibit a high computational burden. Many approaches aimed at reducing complexity exploit fast Fourier transform (FFT) algorithms, which are very efficient for batch data but are not suited for continuously streaming data, typically encountered in software-defined radio (SDR) applications. The aim of this paper is to develop an out-of-tree (OOT) GNU Radio module containing a set of building block functions, aimed at estimating functions, such as the cyclic correlation functions, typically employed for cyclostationary signal analysis. The proposed implementation must be designed so as to ensure high efficiency in processing continuous data streams of complex samples, both in terms of computational load and data storage, in order to be executed on general purpose computers. We tested our implementation by estimating in real-time the second-order cyclic statistics of a Gaussian minimum-shift keying (GMSK) modulated signal, which belongs to a class of signals commonly employed in aeronautical telemetry datalinks.
title CycloDSP: A cyclostationary signal analysis tool for GNU Radio
topic Signal Processing
url https://arxiv.org/abs/2405.16911