More WiFi for everyone: Increasing spectral efficiency in WiFi6 networks using a distributed OBSS/PD mechanism

Authors: ALİ KARAKOÇ, HÜSEYİN BİRKAN YILMAZ, MEHMET ŞÜKRÜ KURAN

Abstract: In this paper, we propose a distributed algorithm that determines effective Overlapping Basic Service Set/Preamble Detection (OBSS/PD) threshold levels in each WiFi6 device to maximize the total throughput by increasing the spectral efficiency. Within WiFi6 standard, OBSS/PD mechanism is introduced to increase the overall efficiency of WiFi networks by tuning the receiver sensitivity as well as the transmission power. In a nutshell, the proposed algorithm, RACEBOT, tunes the hearing (i.e. reception) and speaking (i.e. transmission) parameters of each WiFi device individually for the betterment of the WiFi experience of all WiFi networks in a neighborhood. WiFi experience is not only affected by the spectral efficiency but also by the so-called rate selection algorithms that aims to select the ideal modulation and coding levels according to the ever-changing channel conditions. For higher flexibility, the proposed algorithm works agnostically to the selected rate selection mechanism. Moreover, RACEBOT is a distributed algorithm working independently in all WiFi devices in a given environment. We have conducted extensive simulations in NS3 to evaluate the performance of our proposed algorithm and compare its results with other prominent carrier sensitivity threshold algorithms, both pre-WiFi6 and post-WiFi6. Our results show that RACEBOT outperforms its competitors the RTOT and DSC algorithms in terms of aggregate throughput by 10%-20% in dense networks and 5%-10% in sparse networks considering modern rate selection algorithms such as Thompson.

Keywords: WiFi, 802.11ax, OBSS/PD, spectral efficiency

Full Text: PDF