標題: Stochastic Multiple Channel Sensing Protocol for Cognitive Radio Networks
作者: Hsu, Shao-Kai
Lin, Jia-Shi
Feng, Kai-Ten
電機工程學系
Department of Electrical and Computer Engineering
公開日期: 2011
摘要: A great amount of research has devoted to cognitive radio (CR) in recent years in order to improve spectrum efficiency. In decentralized CR networks, the CR users are expected to be capable of dynamically and opportunistically accessing unused spectrums in primary networks. However, since the spectrum of primary networks is comparatively wide, it is not realistic for the CR users to sense the entire spectrum in practice. Consequently, the partially observable Markov decision process (POMDP) is utilized to provide the CR users with sufficient information in partially observable environments. Moreover, existing POMDP-based protocols exploit techniques of channel aggregation in order to improve the spectrum opportunities and system performance. However, the required time for channel sensing is neglected, which is considered inevitable to result in large sensing overhead and spectrum opportunity loss in realistic environments with increased number of channels. Therefore, in this paper, the stochastic multiple channel sensing (SMCS) protocol is proposed to conduct optimal decision-making based on partially observable channel state information under the consideration of sensing overhead. By adopting the proposed SMCS protocol, the CR user can highly accommodate itself to the rapidly varying environment since the optimal decision-making on multiple channel sensing is dynamically adjusted. Furthermore, the steady-state based SMCS (SMCS-S) scheme with simplified decision-making process is proposed in consideration of implementation complexity. Numerical results illustrate that the proposed SMCS protocol can effectively maximize the aggregated throughput for decentralized CR networks.
URI: http://hdl.handle.net/11536/15180
ISBN: 978-1-61284-254-7
期刊: 2011 IEEE WIRELESS COMMUNICATIONS AND NETWORKING CONFERENCE (WCNC)
起始頁: 227
結束頁: 232
Appears in Collections:Conferences Paper