Impact of Regional Weather on PLC-based Data Transmission in Peninsular Malaysia and Nigeria

Abstract

This paper investigates the potential impact of weather on power line communication (PLC) based systems concerning Throughput. A proposed multi-linear regression-based neural network (MLR-NN) model predicts regional performance of PLC-based data transmission in Malaysia and Nigeria. The paper evaluates the prediction using ITU–T G.9960 which recommends 85–98% Throughput performance for wireline-based technologies like PLC. This paper con- siders related works from US and China for comparison. The regional performance prediction in Peninsular Malaysia is 71.62%, while Nigeria’s regions; Northeast (99.97%), Northwest (79.85%), Northcentral (96.59%), Southeast (81.89%), Southwest (86.32%), and Southsouth (74.46%). Wind analysis in Eastern China showed a high data fitness value (0.995), while in Malaysia (0.985) and Nigeria’s regions (0.980–0.989). Thunderstorms analysis in Southwest US showed lower data fitness (0.630) compared to Malaysia (0.970) and Nigeria’s regions (0.870–0.980). The data fitness (R-square) describes suitability, and strength for reliable ML-based analysis as utilized in this paper.