| Περιγραφή: |
Решается задача адаптивной компенсации дисперсионных искажений в трансионосферных каналах спутниковой связи по данным о среде, получаемых от глобальных навигационных систем. Представлен общий подход, объединяющий в одной системе анализирующее, обучающее и корректирующее программноаппаратные устройства. Приведены аналитические выражения для реализации математической модели адаптивной коррекции частотной дисперсии, оценки времени жизни скорректированного канала. Учтено влияние ошибок определения полного электронного содержания (ПЭС), имеющих стохастический характер, на возможности коррекции дисперсии. Приведены результаты натурных экспериментов, проведённых в ЙошкарОле, позволившие оценить суточные вариации полного электронного содержания, параметры частотной дисперсии, время жизни канала по данным спутниковой системы ГЛОНАСС. Установлено, что присутствующая на практике погрешность оценки ПЭС позволяет расширить полосу связного сигнала только до трёх раз. Introduction. Current trends of developing space communication systems focus on increasing their performance (high noise immunity, channel capacity, low transmitted power). At the same time, there is a need in spreading channel bandwidth. However, to spread the bandwidth of transionospheric channels one have to consider the negative effect of frequency dispersion, that is caused by the dependence of the phase velocity of a wave on its frequency. This results in distortions of signal shape due to the changes in phase relations of the signal components and consequently deteriorates the quality of received data in space communication systems. The aim of the research was to develop the comprehensive approach to adaptive compensation for nonlinear frequency dispersion in wideband space communication systems that employ transionospheric radio channels. Practical significance. Mitigation of dispersion distortions for improving communication system performance and significant spreading transionospheric channel bandwidth. General approach to adaptive compensation. Dispersion corrector is a measuring hardware and software system. Radio engineering system at the receive terminal consists of analyzing, training and correcting devices. Analyzing device performs estimation of the current channel state by diagnosing channel impulse response and transfer function. The important part in creating adaptive compensation device is developing corrector training mode. Contamination of parameters of a transionospheric channel can be attributed to the TEC variations that vary depending on different factors: time of day, season, geomagnetic conditions, solar activity. It means that the distortions will appear in the equalized channel after the certain time period. Therefore, equalizer training requires updating correction functions to compensate FR phase after certain time intervals, also termed to as channel life time. Conclusion. A comprehensive approach to adaptive compensation for frequency dispersion in transionospheric radio channel, that allows to mitigate dispersion distortions of wideband communication signals in case of presence of stochastic error in measuring TEC, was developed. Hardware and software system of dispersion corrector that performs functions of analysis, training and equalization was created. There is presented scientific justification of the mathematical model of the adaptive equalizer. Analytical expressions were derived to estimate TEC with the use of geophysical data on TEC. It was shown that errors in TEC measured by the navigation satellite data give rise to residual dispersion distortions. However, coherence bandwidth may be increased by up to a factor of three. Correction remain actual within the channel life time, that rises with the increase in the operating frequency and decreases with the surge in the channel bandwidth and TEC change rate. The lowest value of the channel life time was 3040 minutes and observed in quiet midlatitude ionosphere. During the periods of disturbances, it drops, therefore the period for data actualization should be approximately 510 minutes. |