Resumen
Interference from other users, due to the long-distance propagation of HF signals, causes spectral congestion, which is considered a primary detrimental effect on HF (3?30 MHz) communication systems. More than a year of HF electric field measurements from a calibrated monopole HF antenna collected by a dedicated measurement system installed in Cyprus have been used to compare neural networks and non-linear regression for the development of single-allocation congestion models in the HF spectrum. These models can be used to advise operators on typical interference spectral congestion levels and assist in the planning and management of frequency usage by assessing the interference level to alleviate spectral congestion in the HF band over Cyprus. The performance of these alternative modeling approaches is discussed in the context of their ability to successfully capture the diurnal variability of HF interference at a certain incident field strength threshold.