DCT-Based OTFS With Reduced PAPR
Nisal Vihan Kalpage, Preety Priya, Yi Hong
Abstract
In this letter, we devise a discrete cosine transform (DCT)-based orthogonal time frequency space (OTFS), offering reduced peak-to-average power ratio (PAPR) and lower complexity without compromising on bit error rate (BER) performance. The proposed system adopts a DCT-based inverse discrete Zak transform (IDZT) to directly convert 2D information symbols to the delay-time domain. Under such a setting, we provide a comprehensive analysis of PAPR. We derive an upper bound on PAPR for DCT-OTFS, and show a reduction of about 20% in the maximum PAPR of traditional OTFS. We also provide a complementary cumulative distribution function (CCDF) analysis of PAPR for this system. Finally, simulations compare the performance of the proposed scheme with orthogonal time sequency multiplexing (OTSM) and OTFS in terms of CCDF of PAPR and BER. We observe that three schemes have similar BER performance, while DCT-OTFS demonstrates an approximate reduction of 0.5 dB in CCDF of PAPR, compared to traditional OTFS and OTSM. We remark that this PAPR reduction is obtained without using any specific PAPR reduction techniques.