Vehicular visible light communications noise analysis and modeling.

Area of Science:

• Electrical Engineering
• Communications Engineering
• Intelligent Transportation Systems

Background:

• Vehicular visible light communications (VVLC) utilizes LED technology for intelligent transportation.
• Noise significantly degrades VVLC system performance.
• Traditional noise models often assume additive white Gaussian noise (AWGN).

Purpose of the Study:

• To analyze time-correlated and white noise components in VVLC channels.
• To develop a more accurate noise model for VVLC systems.
• To evaluate the error performance of VVLC systems under complex noise conditions.

Main Methods:

• Noise analysis using Allan variance for time-series data.
• Development of a motion detection algorithm based on adaptive Gaussian mixture (GM) models.
• Derivation of analytical expressions for error probability in BPSK and QPSK constellations.

Main Results:

• Allan variance analysis identified white noise and random walk in VVLC systems.
• A double Gaussian model was generated for VVLC channel noise.
• Error performance analysis showed significant degradation in the presence of GM noise compared to AWGN.

Conclusions:

• VVLC channel noise exhibits characteristics beyond simple AWGN, including time-correlated components.
• The proposed GM noise model accurately reflects observed noise behavior.
• System performance in VVLC is sensitive to the mixing coefficients of the GM noise distribution, necessitating advanced error mitigation strategies.