Wi-Fi Self-Coexistence in the 6 GHz Band: An ns-3 Evaluation of LPI and SP Usage

Technical Summary: Wi-Fi Self-Coexistence in the 6 GHz Band

Problem Statement
The U.S. Federal Communications Commission (FCC) has authorized unlicensed operations in the 6 GHz band (5.925–7.125 GHz) under four distinct power regimes: Standard Power (SP), Low-Power Indoor (LPI), Geofenced Variable Power (GVP), and Very Low Power (VLP). While these regimes are designed to protect incumbent services, they introduce significant self-coexistence challenges within Wi-Fi networks. Specifically, the heterogeneous transmit power levels—ranging from 14 dBm (VLP) to 36 dBm (SP)—create asymmetric carrier sensing conditions. In a mixed SP and LPI deployment, high-power SP devices may fail to detect lower-power LPI transmissions, while LPI devices aggressively defer to SP activity. This power asymmetry fundamentally alters spatial reuse mechanisms, potentially leading to unfair channel access and severe performance degradation for LPI networks. Prior literature has largely focused on symmetric deployments or Wi-Fi coexistence with cellular technologies, leaving a gap in understanding self-coexistence issues specifically between SP and LPI Wi-Fi 6E networks.

Methodology
To address this gap, the authors developed a system-level simulation testbed using ns-3 with the IEEE 802.11ax (Wi-Fi 6E) model. The study employs a reproducible methodology to evaluate coexistence under two primary scenarios:

1. Indoor LPI–SP: Both APs are located indoors.
2. Indoor–Outdoor LPI–SP: The LPI AP is indoors, while the SP AP is outdoors.

These scenarios are benchmarked against an Indoor LPI–LPI baseline. The simulation parameters include:

• Topology: Two APs (AP0 and AP1) serving five STAs each, with AP separation swept from 40 m to 360 m.
• Traffic: Constant Bit Rate (CBR) downlink UDP flows (1200-byte packets) to ensure a contention-limited regime.
• Configurations: Evaluations cover channel bandwidths of 20, 80, and 160 MHz; Modulation and Coding Schemes (MCS) 0, 5, and 9; and varying OBSS-PD (Overlapping Basic Service Set - Preamble Detection) thresholds.
• Propagation: A hybrid model combining distance-based pathloss and building attenuation (office model) to simulate indoor and outdoor-to-indoor penetration loss.

Key Contributions
The paper makes the following specific contributions:

• Simulation Framework: It constructs a reproducible ns-3 testbed for heterogeneous power regimes, parameterizing sweeps over bandwidth, MCS, load, and OBSS-PD thresholds.
• Quantification of Degradation: It demonstrates that the presence of an indoor SP AP significantly degrades the goodput of an LPI AP compared to an LPI–LPI baseline.
• Bandwidth Analysis: It identifies channel bandwidth as a critical factor in determining the severity of SP-to-LPI impact, noting that degradation is most severe at 20 MHz and partially alleviated at 160 MHz.
• Physical Isolation Effects: It shows that physical blockage (walls) in indoor-outdoor scenarios improves fairness by reducing the airtime gap between SP and LPI APs.
• Spatial Reuse Evaluation: It analyzes the efficacy of BSS coloring and OBSS-PD, finding that aggressive thresholds do not necessarily improve fairness in mixed-regime deployments and may even harm performance.

Results
The evaluation yields four primary findings:

1. Goodput Degradation: LPI APs experience significant goodput loss when coexisting with SP APs. This impact is most pronounced at MCS 9 (high-rate transmission) and 20 MHz bandwidth, where the transmit power difference between SP and LPI is approximately 18 dB. At 160 MHz, the power difference narrows to ~9 dB due to regulatory EIRP caps, resulting in less severe degradation.
2. Distance and Fairness: At inter-AP distances ≤100 m, carrier sensing remains mutual, and fairness is relatively balanced. Beyond 100 m, the power asymmetry causes the LPI AP to defer to the SP AP while the SP AP fails to detect LPI activity, leading to severe goodput loss and reduced Jain's fairness index for the LPI network.
3. Outdoor Impact: In indoor-outdoor scenarios, physical walls provide isolation that improves fairness compared to purely indoor scenarios. However, the outdoor SP AP still suppresses indoor LPI performance, particularly at narrow bandwidths, due to asymmetric sensing.
4. Limitations of OBSS-PD/BSS Coloring: Increasing the OBSS-PD threshold for the LPI AP (to make it more aggressive) does not restore fairness. Instead, it increases exposure to SP transmissions, leading to further goodput degradation. The paper concludes that BSS coloring is insufficient to mitigate SP dominance in these heterogeneous power scenarios.

Significance
The paper asserts that its findings highlight the critical need to understand coexistence behavior among unlicensed devices with varying power constraints, particularly as outdoor SP deployments increase and as regulatory bodies (such as Ofcom in the U.K.) consider sharing the 6 GHz band between Wi-Fi and cellular systems. The study provides a foundational system-level analysis showing that current spatial reuse mechanisms (like BSS coloring) may not be effective in resolving fairness issues caused by power asymmetry in 6 GHz Wi-Fi networks. The simulation framework is presented as extensible for future studies involving VLP devices and denser multi-AP topologies.