【摘要】 在无线通信中,高数据传输率的数字无线系统要求用频谱有效的线性调制方法,但是这些调制方法对功放的非线性又很敏感,会产生频谱扩展、邻近信道干扰和误码率(BER)特性的恶化。本文提出用模糊神经网络(FNN)的算法来实现功放的自适应预失真,以补偿功放的非线性,并仿真了模糊神经网络对功放非线性的补偿以及对误码率特性的改进。结果表明,此方法实现的预失真器具有良好的自适应性和鲁棒性,不需要从一大堆原始数据中进行费时的训练,而可以充分地利用己有的知识和经验;而且,在学习的过程中,采用变结构的神经网络,先粗后细、分组学习,更大大缩短了学习的时间。更多还原

【Abstract】 In digital radio systems, high data rate demands the use of spectrally efficient linear modulation techniques, but these techniques are sensitive to the nonlinearity of high power amplifier (HPA).They give rise to spectral spreading, adjacent channel interference and degraded bit error rates(BER).This paper proposes to implement adaptive predistortion to compensate for the nonlinear nature of HPA with fuzzy neural network(FNN).Simulations show that this method has good adaptability and robustness to changes of HPAs characteristic. In addition, a changing structured neural network is used by group learning. So learning time can be largely reduced.更多还原