A Novel Model for Injecting Error in Probabilistic Gates

Modeling inaccurate gates shows new challenges for electronic design automation (EDA) tools which mainly utilize deterministic approaches in their methods. This paper presents an abstract probabilistic model which depends mainly on the probability of error in the device. We add uniform and normal noise distributions on the probabilistic model for simple gates to assess our proposed model. Also, four different topologies of XOR gates and chain of inverter gates are used to investigate the reliability of simple probabilistic gates. This model shows accurate performance for all the gates by evaluating new metric referred to output percentage error (OPE). Finally, this metric is proved to be as the cumulative distribution function (CDF) for the fitted probability distribution function (PDF) of the injected noise.