For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Study on Soft Decision Based Cooperative Sensing for Cognitive Radio Networks
Hiromasa UCHIYAMA Kenta UMEBAYASHI Takeo FUJII Fumie ONO Kei SAKAGUCHI Yukihiro KAMIYA Yasuo SUZUKI
IEICE TRANSACTIONS on Communications
Publication Date: 2008/01/01
Online ISSN: 1745-1345
Print ISSN: 0916-8516
Type of Manuscript: Special Section PAPER (Special Section on Cognitive Radio and Spectrum Sharing Technology)
Category: Spectrum Sensing
cognitive radio, cooperative sensing, energy detection, cognitive MIMO mesh network,
Full Text: PDF>>
In this paper, we propose a soft decision based cooperative sensing method for cognitive radio (CR) networks for opportunistic frequency usage. To identify unused frequency, CR should exploit sensing technique to detect presence or absence of primary user and use this information to opportunistically provide communication among secondary users while performance of primary user should not be deteriorated by the secondary users. Because of multipath fading or shadowing, the detection of primary users may be significantly difficult. For this problem, cooperative sensing (CS), where gathered observations obtained by multiple secondary users is utilized to achieve higher performance of detection, has been investigated. We design a soft decision based CS analytically and analyze the detector in several situations, i.e., signal model where single-carrier case and multi-carrier case are assumed and two scenarios; in the first scenario, SNR values of secondary users are totally equal and in the second scenario, a certain SNR difference between secondary users is assumed. We present numerical results as follows. The first scenario shows that there is little difference between the signal models in terms of detection performance. The second scenario shows that CS is superior to non-cooperative sensing. In addition, we presents that detection performance of soft decision based CS outperform detection performance of hard decision based CS.