Faster than Nyquist broadcast signaling

Abstract

In this work, we propose using Faster than Nyquist (FTN) signaling to achieve transmission over continuous-time Gaussian broadcast channel. In the FTN broadcasting, different users' messages are completely separated in the coding stage; hence removing the joint encoding that is typically needed in the previously proposed broadcast channel coding schemes. Furthermore, users' data symbols are explicitly and separately transmitted over the channel using the FTN signaling; thus, alleviating the definition of auxiliary random variables used in the broadcast channel coding theorems. This work shows that the FTN broadcast signaling allows achieving the capacity boundaries of the Gaussian broadcast channel. This implies that FTN can be a viable alternative to the standard capacity-achieving techniques considered in discrete-time Gaussian broadcast channel, i.e., superposition coding and dirty paper coding.