ABSTRACT Linear precoding is a means to increase the throughput of the multi spot beam satellite user (forward) downlink by facilitating aggressive frequency reuse compared to a 4‐color reuse scheme. We propose a new way to apply the DVB‐S2X standard in the context of precoding. The approach sidesteps the need to align physical layer frames of simultaneous users on jointly precoded carriers. Hence, physical layer frames can be scheduled more freely by per carrier independent adaptive coding and modulation processes. By introducing the concepts of superframe virtual carrier (SFVC) and dedicated estimation SFVC (ESFVC), a carrier can still be shared by multiple users. In the legacy use of DVB‐S2X for precoding there are syntax elements (non‐precoded pilots) that have dual use: for channel matrix estimation and as demodulation reference for precoded data symbols. We will demonstrate an adverse effect of this legacy approach and provide a remedy. We further sidestep the need for pre‐calibration of differential delays between jointly precoded carriers of large bandwidth in the satellite and gateway RF paths. This is achieved by providing, while staying entirely within the syntax rules of the DVB‐S2X standard, a large non‐precoded field, eventually aligned and quasi‐orthogonal among carriers, however already suitable for estimating the channel matrix and differential time delays even in the presence of initial delay differences exceeding one symbol period. The latter is not the case for methods described in the DVB‐S2X standard's guideline. Finally, we describe how to organize aggressive frequency reuse over large areas served by overlapping spot beams in a scalable way, sticking to small precoding domains and decentralized processing. This contrasts with the concept of a centralized gateway presented elsewhere and sidesteps the need for fine time/frequency alignment of jointly precoded carrier feeds from different RF gateways.
Delaruelle et al. (Wed,) studied this question.