查看文章

In direct sequence code division multiple access (DS/CDMA) wireless networks, each subscriber transmits the symbol modulated by a unique spreading code. The best performance of these systems is achieved when all spreading waveforms are orthogonal. Unfortunately, due to the random timing offsets between users in wireless CDMA networks, the orthogonal property cannot be maintained. To mitigate the effect of time mismatching in the receiver due to the random timing offsets, a complex system is proposed in [1]. In this paper, two CDMA detectors based neural networks are proposed to mitigate the effect of random timing offsets with a less complexity in the detector implementation. The performance of the proposed detectors are obtained and compared with the detectors reported in [1]. The results show the convergence between the performance of the proposed neural network detectors and the detectors reported in [1].