Title :
A Neuro-Genetic Hybrid Algorithm Utilizing Outdoors LOS Optical Wireless Channels
Author :
Yakzan, Adnan ; Green, Roger ; Hines, Evor
Author_Institution :
Sch. of Eng., Univ. of Warwick, Coventry, UK
Abstract :
An Artificial Neural Network (ANN) in the domain of optical wireless communication is presented. The network makes use of a genetic algorithm-based selection and different weather models. The objectives target an attempt to exploit the power of hybrid intelligent systems in the outdoor optical wireless channel and a prediction of a reliable adaptation technique that contributes to the best possible bit-error-rate (BER) under certain weather conditions and link characteristics. The reason of using a GA in this type of selection is that the problem has been coded with a binary search, and took more than 3 hours to evaluate the parameter selection, while the GA approach, under several channel characteristics, achieved the process in tens of seconds.
Keywords :
error statistics; genetic algorithms; meteorology; neural nets; optical communication; optical links; radiocommunication; telecommunication computing; wireless channels; ANN; BER; GA approach; adaptation technique; artificial neural network; binary search; bit-error-rate; channel characteristics; genetic algorithm-based selection; hybrid intelligent systems; link characteristics; neuro-genetic hybrid algorithm; optical wireless communication; outdoors LOS optical wireless channels; parameter selection; weather conditions; weather models; Attenuation; Bit error rate; Genetic algorithms; Optical receivers; Optical transmitters; Wireless communication; BER; artificial neural network; genetic algorithm;
Conference_Titel :
Computational Intelligence, Communication Systems and Networks (CICSyN), 2012 Fourth International Conference on
Conference_Location :
Phuket
Print_ISBN :
978-1-4673-2640-7
DOI :
10.1109/CICSyN.2012.18