An approach based on discrete KarhunenLoeve transformation of the DS/SS signals is proposed to estimate PN sequence in lower S/N ratio DS/SS signals. Characteristics of selforganization and principle components extraction of unsupervised neural networks are exploited adequately. Theoretical analysis and experimental results are provided to show that this approach can work well on the lower S/N ratio input signals. An approach based on discrete Karhunen-Loeve transformation of the DS / SS signals is proposed to estimate PN sequence in lower S / N ratio DS / SS signals. Characteristics of selforganization and principle components extraction of unsupervised neural networks are exploited adequately. Theoretical analysis and experimental results are provided to show that this approach can work well on the lower S / N ratio input signals.