⚡ electrical engineering

OTA Characterization of Dual-User IEEE 802.11be EHT-MU Under Transmit-Chain Imbalance

This paper presents a controlled over-the-air characterization of dual-user IEEE 802.11be EHT-MU transmissions under transmit-chain imbalance, revealing that attenuation primarily causes stream-local degradation for the affected user rather than packet-global failure, with LDPC coding offering a measurable robustness margin over BCC.

Original authors: Mir Lodro, Francesco Raimondo, Geoffrey S. Hilton, Mark A. Beach, Andrew C. M. Austin

Published 2026-05-27

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Mir Lodro, Francesco Raimondo, Geoffrey S. Hilton, Mark A. Beach, Andrew C. M. Austin

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a high-tech radio station (the Access Point) trying to broadcast two different radio shows simultaneously to two different listeners (User 1 and User 2) using the same frequency. This is the new "Wi-Fi 7" standard (IEEE 802.11be), which is designed to be incredibly fast and efficient by sending multiple streams of data at once.

The researchers in this paper wanted to answer a specific question: What happens if one of the radio station's speakers starts to get weaker or "muffled" due to a hardware glitch?

Does the whole broadcast fail, causing both listeners to hear static and lose the signal entirely? Or does the station keep broadcasting the main schedule clearly, while only the listener connected to the weak speaker hears their specific show degrade?

The Experiment: A Controlled "Muffle"

To find out, the team built a "soundproof room" (a shielded RF enclosure) to block out all outside noise. They used advanced software-defined radios to act as the transmitter and the two listeners.

They set up a scenario where:

User 1 and User 2 are listening to the same broadcast.
They deliberately turned down the volume on the second speaker (Transmit Chain 2) step-by-step, creating a "muffled" signal for that specific path.
They measured how well the listeners could understand the show using two main tools:
- EVM (Error Vector Magnitude): Think of this as a "fuzziness meter." It measures how blurry the signal gets before it actually breaks.
- BER (Bit Error Rate): This is the "mistake counter." It counts how many words in the message are garbled or wrong.

The Findings: One Fails, One Survives

The results were very clear and surprising in a good way:

The "Global" Signal Stayed Strong: Even as the second speaker got quieter and quieter, the "common" parts of the broadcast (the station ID, the schedule, the start of the show) remained perfectly clear for both listeners. The radio station didn't lose synchronization; the whole system didn't crash.
User 1 Was Unaffected: The listener connected to the strong speaker heard the show perfectly fine the entire time, even when the other speaker was barely whispering.
User 2 Got "Muffled" First: The listener connected to the weak speaker started to hear their specific show degrade.
- First, the signal got "fuzzy" (EVM got worse).
- Then, suddenly, the message became unreadable (BER shot up), and the listener couldn't understand the show anymore.

The Metaphor: Imagine a teacher talking to two students. If the teacher's microphone for Student B gets unplugged, the teacher can still clearly say, "Class is starting now" (the common signal), and Student A can still hear everything perfectly. However, Student B will eventually stop hearing the teacher's specific instructions. The whole class doesn't go silent; only Student B's experience suffers.

The "Coding" Twist: A Better Safety Net

The researchers then tried a clever trick. They changed the "language" used to send the message to Student B.

Scenario A (BCC): They used a standard, simple language. When the speaker got too quiet (around 30 dB of attenuation), Student B couldn't understand the message anymore.
Scenario B (LDPC): They switched to a more robust, "error-correcting" language (Low Density Parity Check). This language is like having a translator who can guess missing words based on context.

The Result: With the better language (LDPC), Student B could still understand the message even when the speaker was much quieter (up to 35 dB of attenuation). This gave them a 5 dB "safety margin." It's like having a slightly better pair of headphones or a smarter decoder that can handle more static before giving up.

The Bottom Line

The paper proves that when a Wi-Fi 7 system has a hardware imbalance (one antenna is weaker than the other), it doesn't cause a total system failure. Instead, the system gracefully degrades:

The common information stays safe.
The other user stays safe.
Only the specific stream connected to the weak antenna gets worse, eventually failing on its own.

Furthermore, using advanced error-correcting codes (LDPC) can significantly delay this failure, giving the system a much wider range of operation before that specific user loses their connection.

Technical Summary: OTA Characterization of Dual-User IEEE 802.11be EHT-MU Under Transmit-Chain Imbalance

Problem Statement
IEEE 802.11be (Wi-Fi 7) introduces Extremely High Throughput (EHT) features, including enhanced multi-user (MU) transmission, which allows an Access Point (AP) to serve multiple users within a single packet structure. While this improves spectral efficiency, it introduces sensitivity to spatial imbalances, unequal channel conditions, and hardware-chain asymmetries. A critical practical question remains: when an impairment affects a specific transmit dimension (e.g., one AP transmit chain), does it cause a packet-global failure (loss of synchronization or common signaling), or does the degradation remain localized to the specific user stream associated with that impairment? While the PHY and MAC features of 802.11be are documented, there is a lack of controlled Over-the-Air (OTA) assessments regarding dual-user EHT-MU robustness under transmit-chain imbalance.

Methodology
The authors conducted controlled OTA measurements in a shielded RF enclosure using National Instruments (NI) software-defined radios (SDRs).

Setup: An NI USRP-2953R acted as the AP, generating a dual-user, non-OFDMA EHT-MU waveform with two transmit chains. An NI USRP-2942R served as the receiver, implementing synchronized dual-branch recovery to represent two single-antenna users, each receiving one spatial stream.
Impairment Simulation: A calibrated precision variable attenuator was inserted into the second AP transmit chain (TX2). A sweep of attenuation values was applied to create a controlled imbalance between the two transmit chains.
Waveform Configuration: The system utilized a 20 MHz channel bandwidth (CBW20) at 2.48 GHz. Each user was assigned one spatial stream with 1500-byte MSDU payloads. The study compared two coding schemes for User 2: Binary Convolutional Coding (BCC) and Low-Density Parity-Check (LDPC) coding.
Metrics: Performance was evaluated using Bit Error Rate (BER), Error Vector Magnitude (EVM) for EHT-Data, and success probabilities for control fields (U-SIG, EHT-SIG common, and EHT-SIG user) and overall packet recovery.

Key Contributions

Repeatable OTA Methodology: Established a shielded-enclosure measurement framework for dual-user non-OFDMA IEEE 802.11be EHT-MU using NI USRP platforms.
Stream-Dependent Characterization: Provided empirical evidence characterizing how controlled TX2 attenuation leads to stream-dependent degradation rather than packet-global failure.
Field-Wise Analysis: Separated packet-stage signaling recovery from user-specific data recovery, distinguishing between common field robustness and payload reliability.
Coding Robustness Comparison: Quantified the performance margin gained by using LDPC over BCC for the sensitive stream under transmit-chain imbalance.

Results

Stream-Local Degradation: As TX2 attenuation increased, the stream decoded as User 1 remained robust, maintaining a BER floor below $10^{-6}$ across the entire tested range. In contrast, the stream decoded as User 2 exhibited progressive EVM degradation followed by a threshold-like collapse in BER and payload success.
Common Signaling Robustness: Control fields (U-SIG, EHT-SIG) and packet detection remained highly recoverable (success probability near 1.0) even as User 2's payload degraded. This confirms that the impairment mechanism is stream-local at the receiver output, not a packet-global failure.
Coding Gain: Replacing User 2's BCC with LDPC coding delayed the BER and payload-success collapse by approximately 5 dB of TX2 attenuation. Specifically, BCC degradation began around 30 dB attenuation, while LDPC maintained performance until approximately 35 dB.
EVM Distribution: Empirical Cumulative Distribution Function (CDF) analysis of subcarrier-level EVM revealed that the BCC case exhibited a heavier tail of high-EVM observations compared to LDPC, explaining the earlier payload failure.

Significance and Claims
The paper claims that in the measured dual-user non-OFDMA EHT-MU configuration, transmit-chain imbalance does not necessarily result in the loss of the entire packet. Instead, the impairment manifests primarily as degradation of the specific user stream associated with the attenuated chain, while the other stream and common signaling fields remain intact. This distinction is significant for understanding the failure modes of Wi-Fi 7 MU-MIMO systems, suggesting that packet recovery can remain interpretable even when one user's payload becomes unreliable. Furthermore, the study demonstrates a measurable coding-dependent robustness margin, showing that LDPC coding provides a tangible 5 dB advantage in operating range for the sensitive stream under these specific imbalance conditions. The authors note that future work will extend this methodology to wider bandwidths, different MCS values, and OFDMA configurations.

The Experiment: A Controlled "Muffle"

The Findings: One Fails, One Survives

The "Coding" Twist: A Better Safety Net

The Bottom Line

Technical Summary: OTA Characterization of Dual-User IEEE 802.11be EHT-MU Under Transmit-Chain Imbalance

More like this