This project aims to apply the artificial intelligence on Satellite Communications to improve the noise and intereference removal from the signal. The received satellite signal will be first processed by noise and intereference removal from the signal by implementing DFT or ANN-based wavelet processing. The implementation in both cases will be based on FPGAs. After the noise is reduced and the intereference is removed, the demodulation process will be performed on the signal.
A graphical representation of the received satellite signal will be gotten as soon as it is received by a virtual analog-digital converter(ADC) and a processing of noise removal will be applied on that graphical representation either by DFT or by ANNs. The processes of the noise reduction and intereference removal will be done based on the assigned frequency band and frequency values within the band. After the Droppler frequency is removed from the received signal by the processing of DFT, a correction process will be applied to recover the original signal.
There will be two approaches to achieve this correction process: This first approach is arithmetic approach by applying mathematical processing on the received signal and the second approach is by applying the artificial intelligence. The first approach is faster because the second approach of artificial intelligence will be based on recursive operations so that it will consume the resources of the receiver. However, the advantages of artificial intelligence will be to substitute models based on the principle of trial and error as it is the best alternative if it is very complex to substitute the mathematical models.
