ML-based frequency estimation and synchronization of frequency hopping signals

A maximum likelihood (ML)-based algorithm for frequency estimation and synchronization of frequency hopping signals is proposed in this paper. By using a two-hop signal model that incorporates the unknown hop transition time, the likelihood function of the received frequency hopping signal is formul...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on signal processing 2005-02, Vol.53 (2), p.403-410
Main Authors: Ko, C.C., Wanjun Zhi, Chin, F.
Format: Article
Language:English
Subjects:
Algorithms
Applied sciences
Detection, estimation, filtering, equalization, prediction
Estimates
Exact sciences and technology
Frequency estimation
Frequency hopping
frequency hopping communication
Frequency synchronization
Information, signal and communications theory
Iterative algorithms
Iterative methods
Kernels
Mathematical models
Maximum likelihood estimation
Narrowband
Parameter estimation
Pilots
Signal and communications theory
Signal processing algorithms
Signal, noise
Spread spectrum communication
Synchronism
Synchronization
Telecommunications and information theory
Time frequency analysis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A maximum likelihood (ML)-based algorithm for frequency estimation and synchronization of frequency hopping signals is proposed in this paper. By using a two-hop signal model that incorporates the unknown hop transition time, the likelihood function of the received frequency hopping signal is formulated. A new iterative method is then derived to estimate the hopping frequencies and hop transition time. Without using any pilot signal or sync bit, the new algorithm is able to implement synchronization and frequency estimation at the same time. Unlike the time-frequency distribution (TFD) and the wavelet-based algorithms in papers by Barbarossa and Scaglione (1997) and by Khalil and Hippenstiel (1996), the new ML algorithm does not require the selection of a kernel or mother wavelet function. In addition, compared with the TFD-based algorithm, it has a better performance with a lower implementation complexity.
ISSN:1053-587X
1941-0476
DOI:10.1109/TSP.2004.840703