A signal detection method based on auto-correlation for the listen-before-transmit

To avoid collisions of calling from users of the same dell group, the technique of listen-before-transmit (LBT) is employed in in the Third Generation High Frequency Automatic Link Establishment. In this paper a signal detection method based on the transformed sample auto-correlation matrix is proposed for the LBT. This method makes use of the property that the signal is usually pertinent, while the noise is randomly scattered. So it can be adapted in the condition when the power of the noise and the signal is unknown to us. Comparing with several traditional methods, this method performs best in our computer simulations, taking account of the LBT. While the role factors that will improve the performance of the method are discussed and simulated.