Techniques for Short-Term Predictions of Atmospheric Noise Levels

Abstract

Short-term predictions of operational characteristics of ionospheric radio communication such as MUF, atmospheric noise, etc., are likely to lead to more efficient use of the HF spectrum. There are no techniques available for the short-term predictions of atmospheric noise so far. The paper describes two techniques for the purpose viz., autocorrelation technique and source location technique. The former predicts the noise level at a specified hour from the measurement at an earlier hour since atmospheric noise level variations constitute a Markov process. The latter predicts the noise level by predicting the type of the day from the characteristics observed at 16 hrs. Prediction accuracies are discussed, and the possibility of refinements is investigated. Possible system applications are discussed, and a typical calculation is carried out which shows that economies in the radiated power of the order of 10 dB are possible.