Signature-assisted rendezvous in OFDM-based cognitive networks using sub-Nyquist samples

In this paper, the problem of rendezvous in cognitive networks from sub-Nyquist samples is studied and demonstrated. Rendezvous refers to the task of secondary user (SU) sensing and network identification, including possibly multiple simultaneous secondary networks in a certain geographical area. Furthermore, when different networks may deploy different frequency bands, identifying the cognitive network(s) of interest via ordinary sensing receivers and classical Nyquist rate sampling results in substantial hardware complexity and costs for SU devices. In this paper, we exploit the cyclostationary embedded signatures in general OFDM/OFDMA waveform context to efficiently identify the networks using sub-Nyquist samples and joint sparse recovery methods. The proposed method does not require the spectrum itself to be sparse but relies on the sparse nature of spectral correlation function of OFDM(A) waveforms, and is shown through extensive simulations to enable efficient and reliable network identification under low Signal-to-Interference-plus-Noise Ratios (SINRs).