Optimal Joint Routing and Scheduling in Millimeter-Wave Cellular Networks

TL;DR

This paper introduces a novel scheduling and routing framework for millimeter-wave cellular networks that optimizes QoS metrics efficiently, even for large networks, using matching theory and approximation algorithms.

Contribution

It proposes a schedule-oriented optimization method based on matching theory, capable of solving large-scale routing and scheduling problems in mmWave networks.

Findings

Optimal scheduling solutions for networks with over 200 mmBSs.

An approximation algorithm achieving over 80% of the optimal throughput.

The algorithms are extended to multi-RF-chain mmBSs and integrated backhaul/access networks.

Abstract

Millimeter-wave (mmWave) communication is a promising technology to cope with the expected exponential increase in data traffic in 5G networks. mmWave networks typically require a very dense deployment of mmWave base stations (mmBS). To reduce cost and increase flexibility, wireless backhauling is needed to connect the mmBSs. The characteristics of mmWave communication, and specifically its high directional- ity, imply new requirements for efficient routing and scheduling paradigms. We propose an efficient scheduling method, so-called schedule-oriented optimization, based on matching theory that optimizes QoS metrics jointly with routing. It is capable of solving any scheduling problem that can be formulated as a linear program whose variables are link times and QoS metrics. As an example of the schedule-oriented optimization, we show the optimal solution of the maximum throughput fair scheduling (MTFS). Practically, the optimal scheduling can be obtained even for networks with over 200 mmBSs. To further increase the runtime performance, we propose an efficient edge-coloring based approximation algorithm with provable performance bound. It achieves over 80% of the optimal max-min throughput and runs 5 to 100 times faster than the optimal algorithm in practice. Finally, we extend the optimal and approximation algorithms for the cases of multi-RF-chain mmBSs and integrated backhaul and access networks.