Sparse OTFS: Underlay Spectrum Sharing for High Mobility Cognitive Radio Networks

In this paper, we propose a novel modulation scheme named Sparse Orthogonal Time Frequency Space (S-OTFS), which sparsely maps data symbols in the delay-Doppler (DD) domain. The proposed S-OTFS is designed for use as a secondary user (SU) waveform in an underlay cognitive radio (CR) system. While the primary system divides the entire spectrum into subbands and allocates them to primary users (PUs), the S-OTFS spreads the data symbols across the entire frequency-time (FT) domain, effectively reducing interference power levels to the primary system. To mitigate the interference from PUs at the underlay S-OTFS receiver, we propose modified linear minimum mean square error (M-LMMSE) equalizers in both time and FT domains. To this end, we rigorously derive the necessary statistical properties of the PU's interference as the explicit functions of the PU's subband location. We investigate various sparse patterns for the proposed S-OTFS and demonstrate through simulations that the scattered-type sparse patterns exhibit superior performance compared to the clustered-type sparse patterns. Most importantly, the proposed S-OTFS-based CR system with the scattered-type sparse patterns consistently demonstrates substantially improved performance compared to conventional underlay CR systems where multicarrier code division multiple access (MC-CDMA) is employed for a SU, especially in high-speed mobile environments.