Neural model for estimation of external noise power of receiver in wireless communication system

This paper presents MLP (Multilayer Perceptron Network) based neural models of external noise factor of wireless communication system receiver. The proposed models are trained by set of measured results from Recommendation ITU-R P.372. The noise factor calculated by proposed models is used in empirical formula to estimate external noise power. Due to great speed and acceptable accuracy, the estimation of external noise power of receiver by proposed models can be good alternative to manual reading from graphs and use of interpolation formulas.