TABLE OF CONTENTS There are a number of unique aspects of a digital radio implementation which allow a wider choice of options in a receiver design. These options include the use of oversampling to achieve a lower noise floor than the chosen converter resolution would normally allow and the use of undersampling as a method of downconversion. These techniques, together with a range of new mechanisms which can add to both spurious and noise specifications, make the design of a digital receiver somewhat different to its analogue counterpart. This section will cover the major aspects of a digital receiver design, suitable for use in a software defined radio application.
Undersampling is the act of sampling a signal at a sampling rate much lower than one quarter of the Nyquist rate (i.e., much lower than half of the signal frequency). If the signal frequency is, for example, 100 MHz, then the minimum required sample frequency, the Nyquist sample rate, is 200 MHz (although practical converters would require this to be at least 250 MHz). This signal would be undersampled by employing a sample rate of <50 MHz.
Undersampling is an important technique as it effectively performs a frequency mixing function on the input signal, downconverting the signal and, at the same time, performing the required (pseudo-Nyquist) sampling. The signal is converted (aliased) down to baseband or the first Nyquist zone and sampled as if it had originally been a baseband signal.
