Stochastic modeling of microcontroller emission

The conducted emissions from a microcontroller are known to be dependent on the program being run in the core. It is important to know which and how many programs are to be tested to fully evaluate the conducted emission behavior of a microcontroller chip. This paper presents a stochastic estimation-based methodology, for estimating electromagnetic emission levels and emission intervals from a microcontroller, based on the functional blocks activated programmatically. The predicted and actual measurement results show a good match with a confidence level of 76%. Secondly, it is found that for present sample, 8 programs are sufficient to compute an emission interval with 95% confidence level.