信息论准则的频谱感知技术

Spectrum sensing based on akaike information criteria

频谱感知是认知无线电通信的前提,针对传统单天线能量检测算法易受噪声不确定性影响的问题,结合MIMO的空间处理能力,提出一种基于信息论准则的频谱感知算法。首先通过实时侦听信道,获得观测信号的相关矩阵;然后利用AIC准则检测周围环境中是否有主用户存在。该算法无需预先得知噪声的功率信息、无需设置判决门限。仿真结果表明,相比于能量检测算法,该算法有效提高了系统的检测概率,对噪声不确定性具有鲁棒性。

Spectrum sensing is a prerequisite for cognitive radio communication, since the traditional algorithm of energy detec-tion using single antenna is vulnerable to noise, this paper proposed a spectrum sensing algorithm based on Information theory which combines the spatial processing ability of MIMO system. This algorithm first detects the channel in real time to obtain the covariance matrix of signal, and then determines whether the primary user exists by using AIC. The proposed one doesn＇t require the power of noise or a determined threshold. The simulation results indicated that this algorithm increases the detection probability and is robust with uncertainty of noise compared with the algorithm of energy detection.