Blind Detection of DS-SS Signals over Fading Channels Using Negentropy or Kurtosis without any prior Knowledge

A direct sequence spread spectrum (DS-SS) signal uses a large bandwidth so that its power spectral density can be well below the noise level. Hence, detecting the presence of such a signal is a difficult task. In this paper, we propose a new method called negentropy for the detection of DS-SS signal presence, and we compare its performance to that of using kurtosis function. The channel is AWGN plus fast flat or slow non-flat fading. Both methods require no prior knowledge of the spreading sequence used by the transmitter, as well as no need to the channel response of each path and timing of transmitted signals. Both methods measure the non-Gaussianity of the estimated autocorrelation of the received signal. Simulation results show that the negentropy outperforms especially in fading channels. Furthermore, both methods can determine the exact code length of a DS-SS signal.