基于小波神经网络的航空通信盲均衡方法

Blind Equalization Algorithm with Wavelet Neural Network in Aeronautical Communication

针对航空通信信道中,多径干扰带来的信号畸变以及信道记忆效应所产生的码间干扰会严重影响通信质量,且现有的均衡算法由于收敛速度较慢无法适应高速航空通信的需求。提出了一种利用均衡器输出信号构建反馈循环的小波神经网络盲均衡算法。该方法利用反馈循环来消除航空信道的记忆效应,并结合小波神经网络时频局域性的特性来消除航空信道的多径效应和信号畸变所产生的码间干扰。仿真结果表明,该方法可以有效提高盲均衡算法的收敛速度,并使输出的信号星座图更加清晰与紧凑,有利于码间干扰的消除,具有一定的实际应用价值。

According to channel of aeronautical communication,the distorted signal that is generated by multipath interference and the intersymbol interference that is caused by memory of communication channel will seriously affect the equality of communication.But the existing equalization algorithms are hard to satisfy the request of high speed aeronautical communication.A wavelet neural network blind equalization algorithm with feedback revised is proposed.The algorithm utilized the feedback revised model to remove the memory of aeronautical communication,and combining the characteristics of frequency characteristics of the wavelet neural network to abate the signal distortion and multipath interference.The simulation results show that,this method can effectively improve the convergence rate of the blind equalization algorithm,and the signal constellation diagram output is more clear and compact.The algorithm is helpful to eliminate the intersymbol interference and has certain practical application value.