Paper Info
Title
Zero-crossings based spectrum sensing under noise uncertainties
Abstract
In this work, the hypothesis testing problem of spectrum sensing in a cognitive radio is formulated as a Goodness-of-fit test against the general class of noise distributions used in most communications-related applications. A simple, general, and powerful spectrum sensing technique based on the number of weighted zero-crossings in the observations is proposed. For the cases of uniform and exponential weights, an expression for computing the near-optimal detection threshold that meets a given false alarm probability constraint is obtained. The proposed detector is shown to be robust to two commonly encountered types of noise uncertainties, namely, the noise model uncertainty, where the PDF of the noise process is not completely known, and the noise parameter uncertainty, where the parameters associated with the noise PDF are either partially or completely unknown. Simulation results validate our analysis, and illustrate the performance benefits of the proposed technique relative to existing methods, especially in the low SNR regime and in the presence of noise uncertainties.
Year
DOI
Venue
2014
10.1109/NCC.2014.6811308
NCC
Keywords
Field
DocType
cognitive radio,probability,spread spectrum communication,false alarm probability constraint,goodness-of-fit test,hypothesis testing problem,near-optimal detection threshold,noise distributions,noise model uncertainty,noise parameter uncertainty,powerful spectrum sensing technique,weighted zero-crossings,zero-crossings based spectrum sensing,spectrum sensing,goodness-of-fit,noise uncertainties,non-gaussian noise,zero-crossings,goodness of fit
Value noise,Noise floor,Noise (signal processing),Noise measurement,Noise (electronics),Algorithm,Statistics,Quantum noise,Gaussian noise,Mathematics,Gradient noise
Conference
Citations 
PageRank 
References 
3
0.40
13
Authors
3
Name
Order
Citations
PageRank
Gurugopinath, S.1104.65
Chandra R. Murthy254275.73
Chandra Sekhar Seelamantula3487.95