摘要
Anideaofestimatingthedirectsequencespreadspectrum(DSSS)signalpseudo-noise(PN)sequenceispresented.WithouttheapriorityknowledgeabouttheDSSSsignalinthenon-cooperationcondition,weproposeaself-organizingfeaturemap(SOFM)neuralnetworkalgorithmtodetectandidentifythePNsequence.Anon-supervisedlearningalgorithmisproposedaccordingtheKohonenruleinSOFM.TheblindalgorithmcanalsoestimatethePNsequenceinalowsignal-to-noise(SNR)andcomputersimulationdemonstratesthatthealgorithmiseffective.Comparedwiththetraditionalcorrelationalgorithmbasedonslip-correlation,theproposedalgorithm’sbiterrorrate(BER)andcomplexityarelower.
出版日期
2015年01月11日(中国期刊网平台首次上网日期,不代表论文的发表时间)