本文提出了基于最优线性协作的宽带频谱感知方案.通过次级用户之间的协作,为认知网络的频谱感知提供分集,利用融合中心融合多个次级用户的宽带频谱感知数据来获取最优权重,并生成全局判决统计量,最终使用全局阈值完成最后的检测判决.由于所提出的两种协作宽带感知方案或需要精确地估计授权用户的信号强度和噪声方差,或感知性能不足,因而,还提出一种更易实现的方案.理论分析和仿真结果表明,本文所提出的协作感知方案可以有效地提高频谱感知性能,并且性能优于传统等增益合并方案.
This paper proposes two 6ptimal cooperative wideband spectrum sensing schemes for cognitive radios.Multiple cooperative secondary users serve in the cognitive radio network to provide space diversity for spectrum sensing.After the fusion center acquires the optimal fusion weights,a global threshold is utilized to obtain the final global decision.Since the proposed optimal cooperative sensing schemes require precise estimations of primary user signal strengths and the noise variances,an improved sensing algorithm is also proposed to facilitate the practical implementation.Simulations illustrate the proposed schemes significantly improve the spectrum sensing performance and outperform the conventional equal gain combining scheme.The proposed sensing algorithm can effectively implement the optimal cooperative sensing scheme.