Compressive Gray Space Detection for Interweaved Cognitive Radio Systems

We suggest a method to properly detect gray spaces within an active primary system, such as LTE. Here, gray space describes a small fraction of resources within the licensed band, which is not allocated by the primary system due to scheduling decision. By exploiting the power spectral density of the received OFDM signal and the channel from the primary transmitter a distinction between gray space and Primary User activity can be made. Compressed Sensing methods are used to obtain the Power Spectral Density of the primary channel with a low number of measurement samples. Simulation results are presented for hypotheses testing on gray space detection and it is shown that Compressed Sensing outperforms the classical Regularized Least Square algorithm.