Experimental evaluation of multiple signal identification based on cyclostationarity in spectrum sharing scenarios

An important and widely used detection technique is cyclostationarity-based feature detection because the method does not require prior information such as the signal bandwidth or frame format, and time and frequency synchronization are likewise not required. The problem with conventional cyclostationarity-based feature detection is that the detection probability of weak signals degrades if multiple signals with different received-power levels are captured simultaneously. Multiple signal identification has been studied in order to solve such a problem. The iterative detection method suppresses the effects of previously-detected signals in the cyclic auto-correlation domain, and so improves the detection probability of weak signals. In this paper, the multiple signal identification method is evaluated based on testbed experiments. In addition, the modification of the multiple signal identification method in which the effects of other signals are previously-relieved is proposed and applied to the testbed instead of the iterative detection method. The detection performance levels of the conventional and proposed detection methods are evaluated and compared to the results of computer simulations. The results reveal the effectiveness of the proposed detection method in spectrum sharing scenarios.