Adaptive Neuro-Fuzzy model for path loss prediction in the VHF band

Abstract

Path loss prediction models are essential in the planning of wireless systems, particularly in build-up environments. However, the efficacies of the models depend on the local ambient characteristics of the environments. This paper proposed the Neuro-Fuzzy (NF) model for path loss prediction for Ilorin in the VHF band. Received signal strengths along four different routes were measured using NTA Ilorin transmitter which operates at a frequency of 203.25 MHz as a reference. The predictions of the proposed model was compared to Hata, COST 231, Egli and ECC-33 models which are considered standard and widely used empirical path loss models. The Root Mean Square Error (RMSE) was used as a measure of merit for their performances. Across all the routes visited, an average RMSE of 5.253 dB, 9.487 dB, 14.264 dB, 18.696 dB, and 27.890 dB were obtained respectively for the NF, ECC-33, Hata, COST 231 and Egli models. The NF model result is shown to improve the predictions over the estimates obtained when compared with the other models.