Resource allocation for self-backhauled networks with half-duplex small cells

Abstract

Large-scale ad-hoc deployment of small cells or access points is expected to improve the coverage and spectrum reuse. The need for high-cost optical fiber or similar wired connectivity for backhaul can be mitigated by using radio access spectrum for backhaul links. The radio resources are shared among access links and self-backhaul links. This is referred as in-band self-backhauling. In this paper, we propose a solution for the scheduling, resource allocation and flow control problem for networks with self-backhauled half-duplex small cells. We formulate the resource allocation and scheduling problem as an optimization problem, where the objective is to maximize the sum-utility. The logarithm of user equipment (UE) throughputs is used as the utility function. The optimization problem is solved using Karush-Kuhn-Tucker (KKT) conditions. The solution obtained is used to develop a distributed method where the small cells pass on the most parsimonious information to the macro cell to aid it in its scheduling. Based on the intuition from the optimal long-term fraction of resources to be allocated, we propose a solution for the transmission time interval (TTI)-level scheduling problem. Further, we propose a flow-control algorithm between the macro eNodeB (eNB) and the small cell UEs that determines which small cell UEs data needs to be transmitted based on an appropriate feedback on the UE buffer status from the small cell. System-level simulation results are provided to quantify the performance gains of the proposed algorithms for full buffer traffic under different scenarios.