Abstract: In this article, we offer a version of a synthetic neural network-based method for predicting the rate of rainfall 30 seconds in advance. one can use the rainfall rate to calculate the suitable washout adversary-level, such as a virtual inflection plan in advance, to keep the bit error rate (BER) at internal link optimal tiers and enable steady understanding drift at the connection throughout a rain event. This research uses a sample reputation a method that considers past rainfall trends over Durban (29.8587°S, 31.0218°E).When given three nearby historical rain rates , It is observed that the associated prediction model can estimate a rain price for the near future. When errors are evaluated using the root mean square (RMSE) method on the outcomes of our prediction model, it is clear that the resulting errors are within acceptable ranges for specific rain events under exceptional rainfall regimes.

| DOI: 10.17148/IARJSET.2022.9678