Title :
Hybrid neural network/ray tracing model for radiowave penetration into buildings
Author :
Lee, J.W.H. ; Lai, A.K.Y.
Author_Institution :
Dept. of Electron. Eng., Chinese Univ. of Hong Kong, Shatin, Hong Kong
fDate :
9/11/1997 12:00:00 AM
Abstract :
A feedforward multilayer neural network is used to predict the signal strength of building penetration. A ray tracing technique is employed to improve the prediction performance by introducing various separation distances and illustration angles into the modelling setup. The median prediction error is 1.69 dB and the standard deviation is 1.15 dB
Keywords :
feedforward neural nets; multilayer perceptrons; prediction theory; radiowave propagation; ray tracing; building penetration; feedforward multilayer neural network; hybrid neural network/ray tracing model; illustration angles; median prediction error; radiowave penetration; separation distances; signal strength; standard deviation;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19971077
