基于EGARCH过程的电磁频谱占用状态波动特性分析

被引：4

作者：

王磊

苏东林

谢树果

王国玉

机构：

[1] 北京航空航天大学电子信息工程学院

来源：

电子与信息学报 | 2012年 / 34卷 / 11期

关键词：

电磁频谱; 电磁环境; 自回归移动平均; 指数广义自回归; 条件异方差;

D O I：

暂无

中图分类号：

TN98 [];

学科分类号：

摘要：

针对传统频谱占用度自回归移动平均(ARMA)模型由于未考虑序列的条件二阶矩,导致无法准确描述频谱占用状态的非线性时变特性问题,该文提出一种基于指数广义自回归条件异方差(EGARCH)过程的频谱占用状态时间序列建模方法。首先通过对ARMA模型的剩余残差进行条件异方差性检验,表明频谱占用时间序列存在明显的时域"波动集聚"性;其次基于EGARCH过程构建频谱占用度时间序列模型以及对实测数据的分析,表明该模型相较ARMA模型对频谱占用度的拟合与预测精度更高;最后由EGARCH模型参数存在"杠杆效应"系数,表明频谱占用状态变化对电磁环境波动的影响具有非对称性。研究结果表明EGARCH模型能够量化反映频谱占用状态的复杂非线性时变过程。

引用

页码：2767 / 2773

页数：7

共 10 条

[1] Understandthe predictability of wireless spectrum:a large-scale empiricalstudy. Song Cheng-qi,Chen Da-wei,Zhang Qian. IEEE International Conference on Communications . 2010
[2] An autoregressiveapproach for spectrum occupancy modeling and predictionbased on synchronous measurements. Gorcin A,Celebi H,Khalid A Q,et al. IEEE 22thInternational Symposium on Personal,Indoor and MobileRadio Commission . 2011
[3] A quantitative assessmentof wireless spectrum measurements for dynamic spectrumaccess. Pagadarai S,Wyglinski A M. Proceedings of the 4th International Conference onCROW NCOM . 2009
[4] Time series ARIMA model ofspectrum occupancy for cognitive radio. Wang Zhe,Salous S. IET Seminar onCognitive Radio and Software Defined Radios . 2008
[5] Miningspectrum usage data:a large-scale spectrum measurementstudy. Yin Si-xing,Chen Da-wei,Zhang Qian,et al. IEEE Transactions on Mobile Computing . 2011
[6] Time Series Analysis:Univariate and MultivariateMethods. Wei W S. . 2006
[7] Spectrum usageprediction based on high-order Markov model for cognitive radionetworks. LI YANG,DONG YU-NING,ZHANG HUI,et al. CIT 2010:Proceedings of IEEE Conference onComputer and Information Technology . 2010
[8] Binary time series approach to spectrum prediction for cognitive radio. Yarkan S,,Arslan. Vehicular Technology Conference . 2007
[9] Discrete-time spectrumoccupancy model based on Markov chain and duty cyclemodels. López-Benítez M,Casadevall F. IEEE International Symposium on DynamicSpectrum Access Networks . 2011
[10] A spectrumsurveying framework for dynamic spectrum accessnetworks. Datla D,Wyglinski A M,Minden G J. IEEE Transactions on Vehicular Technology . 2009

← 1 →