Sensing of OFDM signals based on time-domain autocorrelation induced by guard subcarriers

Abstract

Cognitive radio (CR) was proposed in wireless communications to improve spectral efficiency based on the concept of opportunistic spectrum access. Because orthogonal frequency division multiplexing (OFDM) has become the major access technique, an efficient spectrum sensing technique targeted at OFDM-based signals is one of the primary requirements of CR networks. To reduce computational complexity, time-domain based approaches are highly desirable; in addition, to improve feasibility, spectrum sensing techniques should rely on as few primary system parameters as possible. In this work, we investigate the time-domain autocorrelation properties of OFDM-based signals that are induced by the presence of guard subcarriers. Accordingly, we design an efficient spectrum sensing algorithm for detecting the presence of OFDM signals. The proposed scheme is robust and flexible, relying neither on time-varying system parameters nor on channel-related parameters.