Sparse Spectrum Detection with Sub-blocks Partition for Cognitive Radio Systems; 子區塊分割於感知無線電系統之稀疏頻譜偵測

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Communication Engineering > Electronic Thesis & Dissertation > Item 987654321/61534

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/61534

Title:	Sparse Spectrum Detection with Sub-blocks Partition for Cognitive Radio Systems;子區塊分割於感知無線電系統之稀疏頻譜偵測
Authors:	尹遜儒;Yin,Xun-Ru
Contributors:	通訊工程學系
Keywords:	感知無線電;壓縮感知;稀疏;子區塊分割;耐奎斯取樣;複合假設檢定;概似機率比;次用戶;Cognitive Radio;Compressive;Sparse;Sub-block Partition;Nyquist sampling;Composite Hypothesis Testing;Likelihood ratio;Unlicensed users
Date:	2013-08-27
Issue Date:	2013-10-08 15:20:05 (UTC+8)
Publisher:	國立中央大學
Abstract:	感知無線電(Cognitive radio)為一項新興通訊技術，藉由偵測未被充分利用的頻譜，我們可以有效改善現有頻譜使用效率。一般而言，頻譜的偵測需滿足耐奎斯取樣(Nyquist sampling)定理，方能達到有效的頻譜偵測效能。因而在寬頻頻譜感知應用上，感知無線電的頻譜偵測運算複雜度將大幅增加，若為降低複雜度而減少頻譜偵測取樣點將造成頻譜偵測的準確度下降。在寬頻感知系統中，主用戶(Licensed users)頻帶的使用往往呈現稀疏分布，壓縮感測(Compressive sensing)技術可以有效克服上述問題，將寬頻訊號做適度壓縮取樣以減低運算複雜度，同時利用訊號在頻域上稀疏特性來有效提升頻譜偵測準確率。於此篇論文中，我們所考慮的頻譜偵測為多子載波偵測問題，且利用稀疏訊號的事前機率與接收到壓縮的訊號，將問題表示成複合假設檢定(Composite Hypothesis Testing)，算出概似機率比(Likelihood ratio)與門檻值(Threshold)比較得到頻譜偵測的結果。為了簡化計算的複雜度，我們將子區塊分割(Sub-block partition)的方法應用到我們頻譜偵測中，子區塊分割的方法為利用偵測少數的子載波數並將剩餘的子載波當成干擾。最後藉由模擬結果來看，我們所提出的法得到之頻譜偵測效能是可靠的，利用此方法次用戶(Unlicensed users)必須在複雜度與頻譜偵測的準確度達到平衡以發揮頻譜偵測最大效能。 The current static spectrum allocation policy is inefficient because some of the outdated licensed bands are often underutilized. Recently, cognitive radio (CR) has been emerged as a promising way to improve the spectrum utilization by sensing the vacant or sparse spectrum of the licensed bands and sharing the spectrum with the unlicensed users. We all know from the Nyquist sampling theorem that the sampling rate should be at least two times faster than the signal bandwidth. However, the increase of the sampling rate will cause high implementation cost, particularly for wideband spectrum sensing. Compressive sensing is an effective remedy to overcome this problem at the sub-Nyquist rate by utilizing the sparse property of those underutilized licensed bands. In this thesis, the occupancy detection problem is formulated as a composite hypothesis testing problem under a Bayesian framework for which the a-prior probability of the sparse signals and the compressive sensing of the received signals are jointly taken into account. The whole considered spectrum is divided into multiple subcarriers, and the sparse spectrum detection is casted as a multi-subcarrier detection problem in time domain. To simplify the computation complexity, we apply the idea of the sub-block partition to our proposed detection method, where a joint detection over a smaller number of multiple subcarriers is performed by using the sparse property and treating the remaining subcarriers as interfering terms. In the simulation result, the detection accuracy has a reliable performance and unlicensed users have to make a trade-off between complexity and detection accuracy.
Appears in Collections:	[Graduate Institute of Communication Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	694	View/Open

社群 sharing

Loading...