Robust Opportunistic Spectrum Access Based on Channel Quality Information in Varying Multi-Channel Networks

Xiaodong Peng,Xiaofeng Zhong,Limin Xiao,Shidong Zhou

DOI: https://doi.org/10.1109/glocom.2013.6831244

2013-01-01

Abstract:This paper considers the problem of cognitive access to the spectrum of primary users in a hierarchical network with multiple time varying channels. We propose a robust policy which is not only based on the occupation and quality situations of the channels, but also considering the inaccuracy of the channel quality estimation. The contribution in this paper is twofold. First we have established a new and more meaningful performance metric to evaluate the secondary user's total throughput among time varying channels, rather than just the channel occupation rate. The optimization objective of cognitive access is designed to maximize the total effective throughput among multiple channels subject to collision constraints for primary users protection. Second we develop a robust algorithm in the case that the parameter of the channel quality distribution is unknown and thus only can be estimated using finite channel samples. The proposed robust algorithm shows excellent effective throughput performance meanwhile meets the collision constraints without perfect knowledge on channel quality information, which makes the robust algorithm more practical and easy to implement in various application scenarios. In order to validate the theoretical analysis of robust algorithm, numerical results are presented to reveal that by introducing the robust access policy, which can take the advantage of access opportunities of the channel in time and frequency domain when it is free and of high quality, the total effective throughput can be improved significantly.