A signal in additive noise is sampled by a receiver with an unstable clock. Maybe the clock has absolute bounds on its frequency drift and sample time aperiodicity, or maybe the clock has some statistical characterization (maybe an Allan variance). Maybe the signal being sampled was also produced by such a system. Can anyone recommend any brief treatments on this situation's signal model, its spectral analysis, and efficient (or typical) signal processing approaches in this domain?
I notice there is a brief Wikipedia article for "Coherence theory (optics)" which is closely related, if not more general.