Fast Prediction of the Quality of Discrete Message Reception for HF Radio 
Channels 
B.G. Barabashov, M.M. Anishin, and O.Y. Pelevin 
 
Modern computer powers combined with the simulation model (developed by the 
authors) of a skywave field open the opportunity to predict HF communication 
reliability on a qualitatively new basis. The paper presents technique of fast 
prediction of average number of errors in discrete messages transmitted over a 
narrowband HF channel. The prediction procedure is implemented as the software 
package which comprises the following blocks: 
–	generator of the transmitted trial digital sequence having arbitrary shift 
keying modulation; 
–	the simulation model of HF signal field; 
–	noise (interference) generator; 
–	model of a decision circuit; 
–	error detection block. 
The main focus is given to description of proposed simplified version of 
deterministic simulation model of a narrowband channel. The model makes it 
possible to account for actual mechanisms of radio wave propagation in 
magneto-active spatially non-uniform ionosphere for specific helio- and 
geophysical conditions, and provides the possibility to work real time. As input 
parameters, the model takes current diagnostics data of the ionosphere which are 
updated in a given step. According to these data, initial ionization 
distribution is being adjusted. The model outputs the average signal voltage at 
a receiver input, and the sequence of impulse response samples. 
The algorithm of predicting the quality of reception assumes the following 
sequence of operations. The trial sequence samples are generated, and then run 
through the model. The resultant output comprises samples of quadrature 
components of the signal, and noise, with signal-to-noise ratio being given. 
Having been processed by the decision circuit, the sequence goes to error 
detection block where it is compared with the initial trial sequence. At its 
final step, the algorithm predicts the current expected number of errors in the 
binary message of given length.