Open Research

Research

The paper, source code, Jupyter notebooks, and references behind the WaveLynk Conditioned Coherence Index framework.

Publication

Predicting Beamforming Instability in Wi-Fi 7 and 6G Systems Using a Conditioned Coherence Framework

IEEE Format · 2025 Allen High School · Allen, TX

Authors: Neha Abin, Sahil Shah, Yajat Parmar

Abstract

This work introduces the Coherence Cliff as a fundamental phenomenon governing beamforming instability in high-frequency MIMO systems. As wireless networks evolve toward Wi-Fi 7 and 6G, Zero-Forcing (ZF) precoding becomes increasingly susceptible to Channel State Information (CSI) aging caused by rapid channel variation.

We propose the Conditioned Coherence Index (CCI), a unified analytical metric that integrates Doppler-induced decorrelation, channel matrix conditioning, and CSI aging to predict the onset of precoding instability. A deterministic switching rule at γ = 0.6 enables proactive transition from ZF to Maximum Ratio Transmission (MRT), preventing catastrophic performance collapse while maintaining high throughput under stable conditions.

Simulation results confirm that CCI increases monotonically with mobility and delay, and that the threshold γ = 0.6 accurately identifies the boundary of the instability region. Hardware measurements on a Wi-Fi 7 testbed validate 30–41% improvements in outage rate and latency relative to fixed ZF strategies.

Download Paper (PDF) GitHub Repository

PDF link active when paper is uploaded to paper/WaveLynk_Paper.pdf.

At a glance

  • Topic: Adaptive beamforming in MU-MIMO
  • Key contribution: CCI metric and γ = 0.6 threshold
  • Framework: Zero-Forcing ↔ MRT switching
  • Validated on: Wi-Fi 7 hardware + Monte Carlo
  • Trials: 100 independent measurements
  • Format: IEEE conference paper format
  • Code: Python (NumPy, SciPy, matplotlib)
CCI = κ(H) · |J₀(2πfDτ)| · SINR/(SINR+α) · e−βτ/Tc

When CCI ≥ γ = 0.6 → switch ZF → MRT

Open Source Code

Python implementation

Full implementation of the CCI framework, Jake's fading channel simulator, ZF/MRT precoders, and the adaptive controller. All local — no external APIs.

src/cci.py

Conditioned Coherence Index

Full CCI computation: doppler_frequency, coherence_time, channel_autocorrelation (J₀ Bessel), condition_number, compute_cci, should_switch_to_mrt.

src/channel_model.py

Rayleigh / Jake's Channel Model

Jake's isotropic fading simulator: generate_channel_matrix, simulate_channel_sequence, add_csi_error. Preserves temporal correlation properties of the Jake's model.

src/beamforming.py

ZF and MRT Precoders

Numerical implementations: zf_precoder (pseudoinverse), mrt_precoder (conjugate transpose), precoding_error, compute_sinr, normalize_precoder.

src/switching.py

WaveLynk Adaptive Controller

WaveLynkController with hysteresis-based switching, event logging, history tracking, and summary statistics. Uses CCI to drive ZF↔MRT transitions.

Quick start

# Clone and install
git clone https://github.com/yajatp/WaveLynk.git
cd WaveLynk
pip install -r requirements.txt

# Run notebooks
cd notebooks
jupyter notebook

# Or import directly
from src.cci import compute_cci, DEFAULT_GAMMA
from src.switching import WaveLynkController

controller = WaveLynkController()
W, mode, cci = controller.select_precoder(H, velocity=1.5, tau=0.005, sinr_db=20.0)
Jupyter Notebooks

Reproducible simulation pipeline

Four notebooks covering theory, paper figures, statistics, and hardware validation.

notebooks/01

CCI Derivation

4 publication-quality plots: Doppler frequency vs. velocity, J₀ autocorrelation with the γ = 0.6 derivation, CCI component breakdown, and matrix conditioning analysis.

View on GitHub →
notebooks/02

Simulation Figures

Reproduces all 3 paper figures: CCI 3D surface vs. velocity and delay, 3D scatter with threshold plane at γ = 0.6, and two-panel stability analysis.

View on GitHub →
notebooks/03

Monte Carlo Sweep

100-trial robustness sweep across randomized velocity, SNR, and delay. Generates SINR box plots, outage rate comparison, and tail risk curves. Exports results CSV.

View on GitHub →
notebooks/04

Hardware Validation

Loads real testbed CSVs (or synthetic placeholders). Generates packet loss vs. signal, latency CDF, throughput by signal category, and theory-vs-measurement comparison.

View on GitHub →
References

Academic citations

  • J. Choi, D. J. Love, and P. Bidigare, "Downlink Training Techniques for FDD Massive MIMO Systems," IEEE Wireless Communications, vol. 21, no. 4, pp. 22–28, Aug. 2014.
  • J. Choi, D. J. Love, and P. Bidigare, "Channel Aging in Massive MIMO Systems," IEEE Wireless Communications Letters, vol. 6, no. 1, pp. 28–31, Feb. 2017.
  • D. Tse and P. Viswanath, Fundamentals of Wireless Communication. Cambridge, U.K.: Cambridge University Press, 2005.
  • T. L. Marzetta, "Noncooperative Cellular Wireless with Unlimited Numbers of Base Station Antennas," IEEE Trans. Wireless Commun., vol. 9, no. 11, pp. 3590–3600, Nov. 2010.
  • H. Q. Ngo, E. G. Larsson, and T. L. Marzetta, "Energy and Spectral Efficiency of Very Large Multiuser MIMO Systems," IEEE Trans. Communications, vol. 61, no. 4, pp. 1436–1449, Apr. 2013.
  • S. Rangan, T. S. Rappaport, and E. Erkip, "Millimeter-Wave Cellular Wireless Networks: Potentials and Challenges," Proc. IEEE, vol. 102, no. 3, pp. 366–385, Mar. 2014.
  • I. F. Akyildiz, J. M. Jornet, and C. Han, "Terahertz Band: Next Frontier for Wireless Communications," Physical Communication, vol. 12, pp. 16–32, Sept. 2014.
  • W. C. Jakes, Microwave Mobile Communications. New York, NY: Wiley, 1974.
  • M. Giordani et al., "A Tutorial on Millimeter Wave Communications for 5G Cellular Networks," IEEE Communications Surveys & Tutorials, vol. 21, no. 1, pp. 173–196, 2019.
  • IEEE Std 802.11be™-2024, "Wireless LAN Medium Access Control and Physical Layer Specifications: Extremely High Throughput (EHT)," 2024.

Try it yourself

The interactive models let you explore the CCI in real time.

Open Live Models → View Results