在接入主用户授权频段之前,认知用户需要检测该频段是否处于空闲状态,以免干扰主用户通信.利用主用户信号和噪声的不同谱相关特性,研究了基于循环谱的频谱检测方法.将主用户非零循环频率上的接收信号循环谱幅度作为检测统计量,给出了判决准则和检测方法.利用主用户信号在不同循环频率下不同程度的循环平稳特征,多个循环频率之间通过加权迭代合作来提高检测结果的可信度；并通过蒙特卡罗仿真方法验证该方法的可行性.仿真结果表明:加权迭代合作可有效实现频谱检测,且检测性能优于等权合作检测；通过合理选择信号采样点数、循环频率个数、迭代次数进行合作检测既可有效提高检测概率,又能保证检测灵敏度.%To avoid interfering with primary user's communications, cognitive user should de tect whether the primary user's licensed spectrum band is idle before accessing it. According to different spectral correlations between the primary user signal and the noise signal, the spec trum detection method based on spectrum correlation function (SCF) is studied. The amplitude of received signal's SCF at non-zero cyclic frequency of primary user signal is set as test statis tic. And the decision rule and the detection method are also given. Taking advantage of differ ent degrees of cyclostationary feature at primary user's different cyclic frequencies, multiple cyclic frequencies are used to improve detection reliability by iterative weighting cooperation. The detection method is verified to be feasible by Monte Carlo simulation. Simulation results show that iterative weighting cooperation is applicable in spectrum detection, and it performs better than the equal weighting cooperation in detection. The detection probability is improved by reasonably choosing signal samples, the number of cyclic frequencies and the number of iter ations , while the detection agility can be ensured.
展开▼
