Fan-beam projection for modulation classification in optical wireless communication systems.

Area of Science:

• Wireless communication technology
• Optical signal processing
• Machine learning for signal identification

Background:

• Increasing bandwidth in wireless connectivity is a key research focus.
• Adaptive modulation enhances communication system efficiency.
• Modulation format identification (MFI) at the receiver is crucial for adaptive modulation, avoiding additional overhead.

Purpose of the Study:

• To develop a novel MFI algorithm for wireless optical communication systems.
• To improve the accuracy and efficiency of identifying modulation formats.
• To enable blind identification of modulation types without prior information.

Main Methods:

• A fan-beam projection technique is used to generate discriminative patterns from constellation diagrams.
• Constellation diagrams representing eight modulation formats (PSK and QAM) are processed as images.
• Deep learning classifiers including AlexNet, VGG16, and VGG19 are employed and compared for MFI.

Main Results:

• The proposed MFI algorithm achieves high classification accuracy, up to 100%.
• Accurate identification is maintained even at low optical signal-to-noise ratios (OSNR) below 8 dB.
• Performance is evaluated across a range of OSNR values from 5 to 30 dB.

Conclusions:

• The fan-beam projection-based MFI algorithm offers a robust solution for wireless optical communication.
• The method demonstrates superior performance compared to related work, especially in challenging low OSNR conditions.
• This approach facilitates efficient and blind modulation format identification, supporting adaptive modulation strategies.