Probability of error analysis of analog-to-digital converters with multiple 8-PSK modulated signals

In modern satellite systems the uplink RF wideband signal after down conversion to IF is input to the analog-to-digital converter (ADC), the output of which is digitally processed for the purpose of channelization and switching of signals. In such systems, a major implementation limitation is the ADC, which needs to operate at a rate at least two times the received signal bandwidth. In the case of a wideband signal composed of many multiplexed signals, the spectral distribution of quantization noise is also of significant interest. Another very important consideration in the ADC performance analysis is the signal clipping effect that occurs whenever the instantaneous input signal amplitude exceeds the maximum linear range of the quantizer. Since clipping cannot be avoided in most practical situations, the signal power to the quantization- plus- clipping- noise-power ratio is of utmost interest. Such a ratio is in general a function of the quantizer load factor that in turn is the ratio of the input signal average power to the maximum peak power at the quantizer output. Even more important for digital communication is the probability of symbol error obtained in the presence of quantization and clipping noise in addition to the receiver thermal noise. The paper presents simulation results on the quantizer analysis when the input signal is composed of a specified number of 8-PSK modulated signals, which have been band limited by square root raised cosine filters. The results are obtained in terms of the probability of symbol error as a function of quantizer load factor.