Open AccessSpectrum Occupancy State Predictor Based on Recurrent Neural Network
Author(s) -
Rong Fan,
Huayan Guo,
Lujie Di,
Ling Xing
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1345/4/042020
Subject(s) - occupancy , recurrent neural network , computer science , cognitive radio , spectrum (functional analysis) , state (computer science) , artificial neural network , artificial intelligence , scarcity , spectrum management , machine learning , algorithm , engineering , telecommunications , wireless , architectural engineering , physics , quantum mechanics , economics , microeconomics
To cope with the scarcity of spectrum resources and to improve the efficiency of spectrum utilization, spectrum occupancy prediction (or named channel state prediction) technology has been addressed into cognitive radio (CR) in recent years. In the paper, we first address the issue how to model the primary user behaviour in CR. And based on the presented behaviour model, a recurrent neural network (RNN) is chosen to design a spectrum occupancy state predictor. By exploiting the learning capacity of recurrent neural network, it can predict the spectrum occupancy state of primary user behaviour. Meanwhile, in order to reveal the proposed predictor’s advantages, the spectrum occupancy state predictor based on RNN and another existed two spectrum occupancy state predictors are also adapted in the simulation section of the paper. Numerical simulation results illustrate the advantages of the proposed method.