A feature extraction and pattern recognition receiver employing wavelet analysis and artificial intelligence for signal detection in diffuse optical wireless communications

• Published in: IEEE wireless communications Vol. 10; no. 2; pp. 64 - 72
• Main Authors: Dickenson, R.J., Ghassemlooy, Z.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.04.2003; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Applied sciences; Artificial intelligence; Channels; Diffusion; Exact sciences and technology; Feature extraction; Optical distortion; Optical receivers; Pattern analysis; Pattern recognition; Radio frequencies; Receivers; Signal analysis; Signal detection; Systems, networks and services of telecommunications; Telecommunications; Telecommunications and information theory; Transmission and modulation (techniques and equipments); Wavelet analysis; Wireless communication; Performance evaluation; Time-frequency analysis; Indoor installation; Wavelet transformation; Noise reduction; Optical telecommunication; Wireless telecommunication; Optical receiver; Pattern recognition; Neural network; Multiresolution analysis; Pattern extraction
• ISSN: 1536-1284; 1558-0687
• DOI: 10.1109/MWC.2003.1196404

Optical wireless diffuse indoor infrared communication systems have as yet large unrealized bandwidths that are not subject to the same regulatory control as radio frequency systems. Usually, well established RF techniques are used to combat channel imperfections for IR implementations. Here, we introduce a novel receiver system based on the multiresolution time-frequency feature extraction capabilities of wavelet analysis, coupled with the well recognized pattern recognition performance of artificial neural networks for mitigating the effects of bandwidth limiting channel-induced distortion.