Detection of anomalies on satellite channels using signal processing and neural network

Intentional and nonintentional interference causes performance degradation to satellite communication links. The authors examine the detection and identification of interference through the use of digital signal processing and the neural network. They determine the sensitivity of the communication waveform to the anomalies, develop a simulation to evaluate the detection and identification performance of the digital signal processing, and develop a neural network to automate the detection process. The authors describe the simulation model, present the results of the sensitivity analysis, and present the detection performance of the digital processing and neural network. Results indicate that the averaged periodogram is capable of identifying tone jammers and adjacent channel interference (ACI) that degrade the bit error rate (BER) performance; and although the detection performance of the neural network developed is promising, it is not at a stage to detect interference with high accuracy.<>