The outstanding problem of semiconductor noise is to account for a power spectrum of which the intensity appears to increase without limit as the frequency is decreased. There is no simple explanation in terms of relaxation times for a spectrum which is proportional to a constant power of frequency over many decades and which is found in a wide range of different materials. But it can be explained if the rate at which carriers leave the conduction band is treated as a queuing problem rather than a relaxation problem. The phenomenon in question is related to the presence of current, which implies that the carriers are swept away from the neighbourhood of the centres from which they originated, and the time to recombination is governed by the time taken to find a vacant centre elsewhere, not by the probability of falling back to the originating centre.
No takes yet. Share an insight, caveat, or question.
David A. Bell (1958) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: