The Performance of PPM using Neural Network and Symbol Decoding for Diffused Indoor Optical Wireless Links

Artificial neural network (ANN) can be an attractive alternative for adaptive equalization especially while channel is nonlinear or non-stationary. Pulse position modulation (PPM) requires the least average optical power compared to other modulation schemes in line-of-sight links but suffer severely in diffused links. The performance of PPM in a diffused channel can be improved by using different equalization techniques. In this work equalization using ANN is proposed and studied. The ANN equalized PPM shows promising results and its performance is comparable to the traditional equalization techniques. The performance can further be enhanced by using 'soft' decision decoding and the simulation results show a 2 dB gain in signal-to-noise.

[1]  Simon Haykin.  Adaptive digital communication receivers , 2000 .

[2]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[3]  S. Haykin,et al.  Adaptive Filter Theory , 1986 .

[4]  Zabih Ghassemlooy,et al.  Wavelet-AI equalization and detection for indoor diffuse infrared wireless systems , 2005, Int. J. Commun. Syst..

[5]  Joseph M. Kahn,et al.  Wireless Infrared Communications , 1994 .

[6]  Joseph M. Kahn,et al.  Modeling of nondirected wireless infrared channels , 1996, Proceedings of ICC/SUPERCOMM '96 - International Conference on Communications.

[7]  John G. Proakis,et al.  Digital Communications , 1983 .

[8]  Zabih Ghassemlooy,et al.  Indoor optical wireless systems employing dual header pulse interval modulation (DH‐PIM) , 2005, Int. J. Commun. Syst..

[9]  Joseph M. Kahn,et al.  Coding and equalization for PPM on wireless infrared channels , 1998, IEEE GLOBECOM 1998 (Cat. NO. 98CH36250).

[10]  Tomoaki Ohtsuki,et al.  Spatial Multiplexing in Optical Wireless MIMO Communications Over Indoor Environment , 2006, IEICE Trans. Commun..

[11]  John R. Barry.  Sequence detection and equalization for pulse-position modulation , 1994, Proceedings of ICC/SUPERCOMM'94 - 1994 International Conference on Communications.

[12]  Martin Fodslette Møller,et al.  A scaled conjugate gradient algorithm for fast supervised learning , 1993, Neural Networks.

[13]  Narasimhan Sundararajan,et al.  A complex valued radial basis function network for equalization of fast time varying channels , 1999, IEEE Trans. Neural Networks.