自适应动量项BP神经网络盲均衡算法

为了消除数字信号在传输过程中产生的码间串扰,使得接收端能够正确解调,对信道畸变进行有效补偿,在基于动量项BP神经网络盲均衡算法的基础上,提出一种能够自适应调节BP神经网络动量项的盲均衡算法.该算法根据盲均衡过程中误差函数的变化情况,自适应调节BP神经网络的动量项,充分发挥动量项在避免网络训练陷于较浅的局部极小点的优势.仿真实验结果表明,该算法在稳定性及收敛性能上均优于固定动量BP神经网络盲均衡算法.

Modified blind equalization algorithm based on back-propagation neural networks with adaptive momentum factor

To infer the transmitted signal from the received signal and to eliminate the InterSymbol interference of the signal, an adaptive momentum factor is introduced for blind equalization based on BP neural network with momentum factor.The momentum factor is modified adaptively according to the values of mean square error.Stimulation shows that this new algorithm is superior to that using constant momentum only, in such aspects as the convergence speed, the steady residual error and the bit-error-rate.