摘要
针对认知无线电系统中的频谱可预测性问题,研究分析了不同观测尺度下频谱状态时间序列的特点,运用递归图技术和递归定量分析方法,从定性和定量两个方面对4种尺度时间序列的可预测性进行了分析和比较,结果表明:随着观测尺度的减小,频谱状态序列的混沌性和随机性增强,可预测性变差;小时频谱占用度序列具有较高的确定度和可预测性,时隙状态序列的随机性强、可预测性弱。所得结论为进一步建立有效的频谱预测模型提供了有益的参考。
The spectrum predictability in cognitive radio(CR) systems is studied. The properties of spec-trum state sequence in different scales are analyzed by state space restructure method. Then the predictabil-ity in qualitative and quantitative aspects is analyzed and compared by Recurrence Plots(RP) and Recur-rence Quantification Analysis(RQA). The RP and RQA results show that the predictability turns to be lessas the scales decrease. And there exist chaos properties in small-scale spectrum state sequence. The con-clusions provide reference for design of spectrum prediction models in future research.
出处
《电讯技术》
北大核心
2015年第2期124-128,共5页
Telecommunication Engineering
基金
河南省科技攻关计划(132102110220)
河南省教育厅重点研究项目(14B510016)~~
关键词
认知无线电
频谱可预测性
递归图
递归定量分析
cognitive radio
spectrum predictability
recurrence plots
recurrence quantification analysis