Implementation and field experiments of cognitive radio system with cooperative spectrum sensing

In cognitive radios, reliable detection of wireless systems is crucial since these radios share the frequency spectrum with other wireless systems. At the same time, it is necessary to avoid interference with the primary systems. In a heterogeneous environment, however, systems with different priorities have different protection requirements and operating parameters. The cognitive radio must guarantee highly reliable communication in the primary systems, but need not necessarily guarantee reliable communication in the secondary systems. Therefore, classification of the signals in a cognitive radio is necessary to fulfill the above mentioned protection requirements. Cyclostationary detection allows for the classification of coexisting cyclostationary signals with different parameters, whereas energy detection cannot be used for distinguishing primary signals from secondary signals. This method has been theoretically investigated, but it has not been applied to an actual field. This paper presents the experimental results obtained from field trials on the use of cyclic-feature-based methods for distributed signal detection and signal classification.