Modeling and Algorithm Design of Coexistence between MLO NSTR Wi-Fi 7 and Legacy Wi-Fi

Abstract

Wi-Fi 7 or equivalently IEEE 802.11be introduces Multi-Link Operation (MLO) to enhance throughput and reduce latency by enabling simultaneous transmission and reception (STR) over multiple frequency channels. This is realized through Multi-Link Devices (MLDs), which consist of multiple affiliated stations each with its own lower MAC, controlled by a common upper MAC. While AP MLDs generally support STR, non-AP MLDs face challenges due to IDC (In-Device Coexistence) interference from closely placed transceivers. To mitigate this, non-AP MLDs adopt non-STR (NSTR), requiring start- time and/or end-time alignment across links. To fully exploit MLO’s potential, however, it is essential to examine how MLDs would coexist with legacy Single-Link Devices (SLDs). While several simulation studies have evaluated coexistence between MLDs and SLDs, such an ap- proach inherently depends on specific topologies and parameters set up for simulations, often requiring significant time until obtaining meaningful results. In contrast, analytical models provide closed-form expressions that clarify how design parameters affect performance and offer insights beyond simulation. However, only a few analytical studies exist on MLO, and most failed to capture standards-compliant MLO behaviors rigorously. These limitations hinder accurate modeling and design of effective MLO mechanisms. To address these gaps, this dissertation presents the first analytical framework that correctly mod- els AP MLDs and non-AP MLDs in coexistence with legacy devices, fully compliant with the IEEE 802.11be standard. It constructs Markov chain (MC) models that preserve per-STA backoff behavior and incorporate mandatory alignment rules. First, we propose a novel MC model for AP MLDs sup- porting STR operation, to illustrate how MLO behaviors can be incorporated into the MC framework serving as an intuitive foundation. The model is then extended to cover non-AP MLDs under NSTR constraints, through which detailed analyses are conducted for both AP MLDs and non-AP MLDs. As a result, this dissertation will derive key performance metrics in their closed-forms such as transmis- sion probabilities, collision probabilities, and per-device per-link throughput, which are then validated through performance evaluations using our custom-developed ns-3 simulator for Wi-Fi 7 MLO. Finally, building on the proposed analytical models, this dissertation identifies throughput-fairness related issues under conventional channel access schemes. To address them, it proposes dynamic chan- nel access algorithms tailored to MLO environments for achieving balanced performance across various device types. Through such analyses and simulations, the dissertation demonstrates that the proposed models and algorithms successfully capture MLO behaviors and can improve the coexistence perfor- mance.