Enabling the Internet of Things With Wi-Fi Halow—Performance Evaluation of the Restricted Access Window

Evgeny Khorov,Mischa Dohler,Andrey Lyakhov,Ruslan Yusupov,Massimo Condoluci,Ian Akyildiz,Alexander Krotov

doi:10.1109/access.2019.2939760

Abstract

IEEE 802.11ah, a new amendment to the Wi-Fi standard, adapts Wi-Fi networks to the emerging Internet of Things (IoT). A key component of .11ah is the Restricted Access Window (RAW), a new channel access mechanism, which reduces contention when even thousands of IoT devices operate in the same area by assigning them different channel times. This paper shows that existing studies incorrectly understand the RAW behavior, oversimplify its modeling and thereby overestimate the real system throughput in several times, especially for short durations of the reserved RAW slots. The core contribution of this paper is a new mathematical model based on a completely different approach, which yields more accurate results and thereby enables better IoT system dimensioning. The developed model is suitable for many scenarios typical for IoT. It allows finding RAW parameters that optimize system performance in terms of throughput, power consumption, and packet loss ratio. The proposed solution is can be used for various traffic patterns: when each device transmits a single packet, a batch of packets of random size, or it has full-buffer traffic.

Highlights

Wi-Fi was initially designed to provide broadband wireless Internet access for a small number of devices
We evaluate other performance indices such as packet loss ratio and energy consumption, which are very important for Wi-Fi networks [20], [21], especially for Internet of Things (IoT) scenarios
In this paper, we have developed a mathematical model, which — in contrast to prior adaptations of Bianchi’s approach originally designed for traditional Wi-Fi networks — accurately estimates the performance indices of an IEEE 802.11ah network with Restricted Access Window while previous models, like Group-Synchronized Distributed Coordination Function (GS-DCF), may severely overestimate the throughput, up to ten times underestimate the power consumption, and up to 20% underestimate PLR, as we demonstrate with numerical results

Summary

INTRODUCTION

Wi-Fi was initially designed to provide broadband wireless Internet access for a small number of devices. They assume that a STA suspends its backoff counter between RAW slots. Such an assumption allows them to model transmission as a steady-state process, to the well-known Bianchi model. We apply the model to analyze the transmission process in the scenario, when several STAs use the RAW mechanism to transmit data In this scenario, the model allows finding such RAW parameters (the number of RAW slots and their duration) that maximize the throughput and guarantee packet delivery in time within a given delay budget.

PROBLEM STATEMENT

INFLUENCE OF THE NUMBER OF GROUPS

CONCLUSION