摘要
为满足自编码扩频系统中信源序列随机化的要求,对信源随机化方法进行了研究,使用了一种自适应滤波器,利用其前向预测功能从信源序列中提取自编码扩频序列。该滤波器是在线性自适应滤波器中引入非线性神经元网络层构成,采用了一种新的变步长的更新算法,并简化了滤波器各项参数的更新,实现了一种基于自适应滤波器的信源随机化方法。仿真实验结果表明,使用自适应滤波器提取的自编码扩频序列完全满足自编码扩频通信的要求,系统误码率有所改善。
In order to meet the needs of the randomized application of self-encoded spread spectrum source sequences,the methods of randomized are studied,and an adaptive filter is used to extract spreading sequences from source using its forward prediction.This filter is constituted by introducing a nonlinear ANN layer into the traditional linear adaptive filter,a new variable step size updating algorithm is used,and the update of filter parameters are simplified.An adaptive whitening filter is successfully realized and used to randomize the source sequences of self-encoded spread spectrum.The results of simulation testify that the spreading sequence extracted from source using the adaptive filter fully meets the system requirements,the BER of SESS is improved.
出处
《计算机工程与设计》
CSCD
北大核心
2012年第2期460-463,560,共5页
Computer Engineering and Design
关键词
自适应滤波器
神经网络
自编码扩频
变步长
误码率
adaptive filter
artificial neural net
self-encoded spread spectrum
variable step size
bit error rate
