On Throughput Maximization for a Multi-hop MIMO Network

There has been a growing interest to employ the so-called degree-of-freedom (DoF) based models to study multihop MIMO networks. Existing DoF-based models differ in their interference cancelation (IC) behavior and suffer from either loss of solution space or possible infeasible solutions. Recently, a DoF model based on a novel node-ordering concept was proposed to overcome the limitations of the exiting DoF models. In this paper, we apply this new DoF model to study a throughput maximization problem in a multi-hop network. The problem formulation jointly considers half duplex, node ordering, DoF consumption constraints and flow routing and is in the form of a mixed integer linear program (MILP). Our main contribution is the development of an efficient polynomial time algorithm that offers a competitive solution to the MILP through a series of linear programs (LPs). The key idea in the algorithm is to explore (i) the impact of node ordering on DoF consumption for IC at a node, and (ii) route diversity in the network while ensuring DoF constraints are satisfied at each node throughout the iterations. Simulation results show that our solutions by the proposed algorithm are competitive and feasible.