14-06-2012, 03:41 PM
A Neural Network Model for the Critical Frequency of
the F2 Ionospheric Layer over Cyprus
A Neural Network Model for the Critical Frequency.pdf (Size: 345.6 KB / Downloads: 36)
Abstract
This paper presents the application of Neural Networks for the prediction
of the critical frequency foF2 of the ionospheric F2 layer over Cyprus.
This ionospheric characteristic (foF2) constitutes the most important parameter
in HF (High Frequency) communications since it is used to derive the optimum
operating frequency in HF links.
Introduction
Skywave HF communiations utilize the ability of the ionosphere to reflect waves up
to 30 MHz to achieve medium to long-distance communication links with a minimum
of infrastructure (figure 1). The ionosphere is defined as a region of the earth's upper
atmosphere where sufficient ionisation can exist to affect the propagation of radio
waves in the frequency range 1 to 30 MHz. It ranges in height above the surface of
the earth from approximately 50 km to 600 km. The influence of this region on radio
waves is accredited to the presence of free electrons.
Characteristics of the F2 Layer Critical Frequency
Measurements of foF2 are conducted by ionosondes which are special types of radar
used for monitoring the electron density at various heights in the ionosphere. Their
operation is based on a transmitter sweeping through the HF frequency range transmitting
short pulses. These pulses are reflected at various layers of the ionosphere,
and their echoes are received by the receiver and analyzed to infer the ionospheric
plasma frequency at each height. The maximum frequency at which an echo is received
is called the critical frequency of the corresponding layer. Since the F2 layer is
the most highly ionized ionosperic layer its critical frequency foF2 is the highest frequency
that can be reflected by the ionosphere.
Experiments and Results
A Neural Network (NN) was trained to predict the foF2 value based on sinhour,
coshour, sinday, cosday and R (modeled sunspot number) model parameters. The
33149 values of the dataset recorded between 1987 and 1997 were used for training
the NN, while the 3249 values of the more recent dataset recorded from 18.09.08 until
16.04.09 were used for testing the performance of the trained NN. The training set
was sparse to a certain degree in the sense that many days had missing foF2 hourly
values and this did not allow the dataset to be approached as a time-series.
Conclusions and Future Work
In this paper we have presented the development of a neural network model for the
long-term prediction of critical frequency of the F2 ionospheric layer (foF2) above
Cyprus. The model has been developed based on a data set obtained during a period of
ten years and tested on a dataset of seven months that was recently obtained. The model
has produced a good approximation of the different time-scales in the variability of the
modeled parameter.