Autocorrelation Based Detection of Multicarrier Signals with Periodic Power Boosting

Given the prevalence of multicarrier waveforms in wireless systems, the sensing of such signals is of high importance for cognitive radio applications. In this work, we exploit periodic power boosting of subcarriers in multicarrier signals which is applied e.g. in LTE to improve the performance of sensing compared to state-of-the-art schemes under practical scenarios where noise uncertainties are present. First, we establish an analytical connection between the subcarrier power distribution and the autocorrelation function of a general multicarrier signal. Second, using this result, we demonstrate that periodic power boosting causes strong peaks in the magnitude of the autocorrelation function. The obtained analytical results match well with over-the-air measurement results corresponding to an orthogonal frequency division multiplexing (OFDM) signal of an LTE transmission. Here, the measurements indicate the presence of a boosting with a period of three subcarriers in the LTE signal which arises due to the cell-specific reference signals. Subsequently, a sensing algorithm exploiting the special features of the autocorrelation function resulting from periodic power boosting is developed which outperforms energy detection under realistic noise uncertainties and a state-of-the-art method that relies only on the cyclic prefix.