Abstract:

We introduce the relationship between excess noise in Optoelectronic Coupled Devices (OCDs) and their interior defects and explain how low-frequency noise can be used to estimate their reliability. Using concepts from the biological immune system and its process of identifying invaders, we present a system for estimation of the reliability of OCDs. The system has expressions for the antigen (excess noise), lymphocyte (criterion) and the role of the lymphocyte eliminating unreliable devices. A genetic algorithm was used to estimate the components parameters of the noise spectrum for estimating the reliability of OCDs. The experimental results demonstrated that this method is reliable, adaptable and practical.

Key words: low-frequency noise, artificial immunity, genetic algorithm, device reliability