Channel Estimation and Turbo Equalization for Coded OTFS and OFDM: A Comparison

In this letter, we study joint channel estimation and turbo equalization for coded orthogonal time frequency space (OTFS) and orthogonal frequency division multiplexing (OFDM). To the best of our knowledge, this is the first study that compares performance of OTFS and OFDM coded systems using the state-of-the-art message passing (MP) and soft minimum mean square error (MMSE) equalizers, under imperfect channel estimation. We show that the commonly used threshold-based channel estimator incurs noticeable performance loss for OTFS. Hence, we propose a basis expansion model (BEM) channel estimation technique for OTFS, which improves performance of the threshold-based channel estimator and yields a superior performance to OFDM when using small modulations of BPSK and QPSK. Our work also reveals that OTFS can perform inferior to OFDM as the modulation order increases. This is due to the increased 2D detection complexity of OTFS.