Downlink scheduling for users in cell boundaries

Abstract

In this thesis, we consider a problem of optimally scheduling users in cell boundaries, where the users suffer from the interference from neighboring base stations. Under such circumstances, we propose new scheduling schemes that consider interference and encoding/decoing through a cooperation between two base stations. We use some of the best known achievable regions from information theory to schedule users in such interference environments. We extend the conventional $2 \times 2$ interference channel (IC) with 2 transmitters and 2 receivers to $2 \times 2n$ ICs with 2 transmitters and 2n receivers. In the proposed scheduling, we assume that the receiver uses a combination of two detection schemes, single user detection (SUD) and multi-user detection (MUD), and that the transmitter encodes messages in different coding schemes depending on the channel conditions. We consider three scheduling schemes and compare the performance. Among three schemes, two schemes that we propose are the scheduling schemes through full cooperation and partial cooperation and the last scheme is the conventional scheme through no cooperation. We show the improved performance by the four proposed schemes, compared with the performance by the conventional scheme. We show that the proposed schemes achieve more throughput, compared with the throughput by scheduling scheme with SUD only, MUD only and time division multiplexing (TDM) only.