Scheduling strategies for multi-antenna communications and dual connectivity in wireless networks

Abstract

Future generations of mobile networks are envisioned to support a wide variety of use cases, to which a plurality of strict quality of service (QoS) requirements need to be fulfilled. In order to meet such requirements, fifth generation (5G) and beyond mobile networks are expected to rely on, among other techniques, multi-antenna communications and the deployment of small cells. To fully exploit the potential benefits of these techniques, the design of efficient scheduling solutions are paramount since these techniques largely influence the overall performance of the system. In this context, this thesis deals with the design of efficient scheduling strategies for the multiple input multiple output (MIMO) interference broadcast channel (IBC) and for dual connectivity (DC) networks. The MIMO IBC is a general model for downlink communication in which a plurality of multi-antenna transmitters wish to simultaneously send data to the respective intended multiantenna receivers. The main challenge in the MIMO IBC is designing linear transceivers that maximize the system throughput while fulfilling per-user QoS requirements or achieving a level of fairness among the users. In this context, based on optimization theory, this thesis designs centralized, semi-distributed and distributed algorithms for rate-constrained sum-rate maximization in the MIMO IBC. In DC networks, users can be simultaneously connected to more than one base station. In this context, we address scheduling aspects of flow control algorithms that command the data split among the multiple connections of a given user. The proposed solution is based on the utility theory and focuses on maximizing the user satisfaction in the system. Furthermore, we also propose a novel and practical medium access control (MAC) scheduler architecture and corresponding scheduling algorithm for future multi-connectivity networks, where users can even have more than two simultaneous connections. The solutions developed in this thesis are focused on the enhancement of the per-user QoS and are amenable to practical implementations.