Single-Frequency Network Terrestrial Broadcasting with 5GNR Numerology Using Recurrent Neural Network

We explore the feasibility of Terrestrial Broadcasting in a Single-Frequency Network (SFN) with standard 5G New Radio (5GNR) numerology designed for uni-cast transmission. Instead classical OFDM symbol-by-symbol detector scheme or more complex equalization technique, we Recurrent-Neural-Network (RNN)-based that replaces channel estimation and blocks. The RNN is bidirectional Long Short-Term Memory (bi-LSTM) computes log-likelihood ratios delivered to LDPC decoder starting from received symbols affected by strong intersymbol/intercarrier interference (ISI/ICI) on time-varying channels. To simplify receiver reduce system overhead, pilot data signals our proposed are superimposed instead interspersed. describe parameter optimization provide end-to-end simulation results, comparing them those system, where waveform specifically Broadcasting. show outperforms receivers, especially challenging scenarios associated large intersite distance mobility. also evidence robustness receiver, showing an trained single signal-to-noise ratio user velocity performs efficiently range different velocities.