Fractional Delay-Doppler Channel Estimation in OTFS with Sparse Superimposed Pilots using RNNs

In this paper, we consider the problem of delay-Doppler (DD) channel estimation in orthogonal time frequency space (OTFS) modulation with fractional delays and Dopplers. Exclusive use of DD bins in a frame for pilot symbols causes rate loss. Superimposing pilot symbols over data symbols avoids this rate loss. Our contributions in this paper are two-fold. 1) We propose a sparse superimposed pilot (SSP) scheme where pilot and data symbols are superimposed in a few bins and the remaining bins carry data symbols only. This scheme offers the benefit of better inter-symbol leakage profile in a frame, while retaining full rate. 2) For the SSP scheme, we propose a recurrent neural network based learning architecture (referred to as SSPNet) trained to provide accurate channel estimates overcoming the leakage effects in channels with fractional DD. Simulation results show that the proposed SSP scheme along with fractional DD channel estimation using the proposed SSPNet performs better than a fully superimposed pilot scheme.