Theoretical basis and implementational challenges of sampling with internal filtering

Conventional sampling based on traditional interpretation of the sampling theorem has inherent problems. Most of them are related to the necessity to accumulate signal energy in the capacitors of track-and-hold amplifiers (THAs) during a small fraction of the sampling interval. This limits the dynamic range and attainable bandwidth of software defined receivers (SDRs) and makes digitization close to the antenna impossible. Antialiasing filters used by conventional sampling limit adaptivity, reconfigurability, and scale of integration of SDRs. It is shown in this paper that development of sampling methods inevitably leads to the transition from the traditional interpretation of the sampling theorem to its new interpretation. It is also proven that the new interpretation directly follows from the sampling theorem. Novel sampling with internal antialiasing filtering based on the new interpretation accumulates the signal energy over several sampling intervals for each sample. This enables radical improvement of the SDR performance. The paper provides qualitative and quantitative estimation of advantages of the novel sampling for SDRs. Reconstruction of analog signals based on the new interpretation of the sampling theorem is also briefly discussed. An original approach to the analysis of the influence of nonideal integration on the transfer function of novel sampling circuits (NSCs) is presented.