| 摘要: |
| 针对混合扩频信号多是基于直接序列扩频和跳频扩频相结合的现状，提出了基
于直接序列和线性调频相结合的混合扩频信号快速参数估计与识别算法。通过建立基于频域
矩峰度系数的神经网络，将两类信号识别开来, 信噪比为-2 dB时识别概率均达到95%以
上
。通过倒序共轭卷积和分段截取后快速解线调，实现了复合信号各参数的快速估计，信噪比
为
-1 dB时码速率估计精度均较高。 |
| 关键词: 混合扩频 参数估计 调制识别 神经网络 |
| DOI： |
|
| 基金项目:电子科技大学成都学院科技基金项目 |
|
| Fast parameter estimation and recognition approach for hybrid spread spectrum signals |
| ZHANG Jia-fen,MOU Fei-yan,LEI Xiang |
| (Chengdu College,University of Electronic Science and Technology of China,C
hengdu 611731,China) |
| Abstract: |
| To improve the status of hybrid spread spectrum signal mainly based on
direct sequence spread spectrum(DSSS) signal and frequency hopping(FH) spread sp
ectrum sig
nal, a fast approach is presented for DSSS and chirp
spread spectrum. By establishing coefficient of frequency domain moment peak ba
sed neural network, two types of signals are identified. The recognition probabi
lity is above 95% when SNR(Signal-to-Noise Ratio) is close to -3 dB. By reve
rse-order conjugation c
onvolution-fast dechirp, the code rate and initial frequency and slope of convol
utional compound signal can be estimated.Code rate estimation accuracy is highe
r when SNR is close to -1 dB. |
| Key words: hybrid spread spectrum parameter estimation modulation recognition neural netw
ork |
《电讯技术》编辑部 