基于信噪比软信息的协同频谱感知算法

Cooperative Spectrum Sensing Data Fusion Method Based on Signal Noise Ratio Soft-Information

  • 摘要: 针对协同频谱感知方法中假定各基站信噪比近似相同,使信噪比差异较大时检测概率性能无法得以明显提升的问题,本文提出一种在OFDM系统下,结合信噪比信息进行融合判决的协同频谱感知算法。该方法首先利用OFDM系统的循环前缀提出一种适合认知无线电系统的盲信噪比估计方法;然后,利用检测概率在信噪比-5dB~-13dB的区间内近似线性变化的特性,依据估计所得信噪比值对检测概率曲线进行软信息量化,作为各基站判决信息的置信度,从而使权重分配更加合理化,同时,相比于判决融合方法,并不明显提高传输带宽。仿真结果表明,相比与传统方法,此方法能有效利用信噪比信息,在不同信噪比环境下均有较优性能,尤其在各基站信噪比差异较大时,其性能已接近数据判决精度。

     

    Abstract: Aim to the problem that the signal noise ratio (SNR) of each user is assumed to nearly the same, so that the cooperative performance can not be improved remarkable in the larger difference of SNR environments, based on SNR information, a novel cooperative spectrum sensing scheme for OFDM system is proposed. By using the characteristics of cyclic prefix of OFDM signal, the novel blind SNR estimation is proposed without any information of modulation or synchronization for the cognitive radio system. Moreover, by using the characteristics of detection probability curve which is close to the linear between -13dB and -5dB, and according to the estimated SNR, the detection probability curve can be quantized into the soft-information as decision believe degree information in order to attain more reasonable weight in fusion scheme. Computer simulation shows that the proposed scheme can utilize the SNR information effectively. And comparing with the conventional scheme, the performance of proposed scheme is better in various situations. Especially in the larger difference of SNR environments, the performance of proposed scheme is closer to data decision scheme.

     

/

返回文章
返回