Optimal Distributed Scheduling in Wireless Networks under the SINR interference model

TL;DR

This paper introduces the first distributed, message-passing-free scheduling algorithms that are optimal under the realistic SINR interference model, combining power allocation and randomization for wireless network efficiency.

Contribution

It presents novel distributed algorithms that achieve throughput-optimality under the SINR interference model without message passing, a significant advancement over prior graph-based models.

Findings

Algorithms are rate-optimal matching centralized schemes.

Algorithms achieve throughput-optimality in SINR-based interference.

No message passing required for the proposed schemes.

Abstract

Radio resource sharing mechanisms are key to ensuring good performance in wireless networks. In their seminal paper \cite{tassiulas1}, Tassiulas and Ephremides introduced the Maximum Weighted Scheduling algorithm, and proved its throughput-optimality. Since then, there have been extensive research efforts to devise distributed implementations of this algorithm. Recently, distributed adaptive CSMA scheduling schemes \cite{jiang08} have been proposed and shown to be optimal, without the need of message passing among transmitters. However their analysis relies on the assumption that interference can be accurately modelled by a simple interference graph. In this paper, we consider the more realistic and challenging SINR interference model. We present {\it the first distributed scheduling algorithms that (i) are optimal under the SINR interference model, and (ii) that do not require any message passing}. They are based on a combination of a simple and efficient power allocation strategy referred to as {\it Power Packing} and randomization techniques. We first devise algorithms that are rate-optimal in the sense that they perform as well as the best centralized scheduling schemes in scenarios where each transmitter is aware of the rate at which it should send packets to the corresponding receiver. We then extend these algorithms so that they reach throughput-optimality.