Two-Layered Superposition of Broadcast/Multicast and Unicast Signals in Multiuser OFDMA Systems

We study optimal delivery strategies of one common and K independent messages from a source to multiple users in wireless environments. In particular, two-layered superposition of broadcast/multicast and unicast signals is considered in a downlink multiuser OFDMA system. In the literature and industry, the two-layer superposition is often considered as a pragmatic approach to make a compromise between the simple but suboptimal orthogonal multiplexing (OM) and the optimal but complex fully-layered non-orthogonal multiplexing. In this work, we show that only two-layers are necessary to achieve the maximum sum-rate when the common message has higher priority than the K individual unicast messages, and OM cannot be sum-rate optimal in general. We develop an algorithm that finds the optimal power allocation over the two-layers and across the OFDMA radio resources in static channels and a class of fading channels. Two main use-cases are considered: i) Multicast and unicast multiplexing when K users with uplink capabilities request both common and independent messages, and ii) broadcast and unicast multiplexin when the common message targets receive-only devices and K users with uplink capabilities additionally request independent messages. Finally, we develop a transceiver design for broadcast/multicast and unicast superposition transmission based on LTE-A-Pro physical layer and show with numerical evaluations in mobile environments with multipath propagation that the capacity improvements can be translated into significant practical performance gains compared to the orthogonal schemes in the 3GPP specifications. The impact of real channel estimation is also analyzed and it is shown that significant gains in terms of spectral efficiency or increased coverage area are still available even in the presence of estimation errors and imperfect interference cancellation for the two-layered superposition system.