How to best incorporate sensor variability in simulation?

Hi,

First of all thanks for your response in advance. Here is my question:

I have a current sensor spec with typical (mean) value of 500mA with min and max of -/+ 1500 mA. I also have ideal sensor inputs. How do I best incorporate this variability for my simulation purpose? which simulink block to use? random no, uniform random no, or band limited white noise? What I am looking for is the impact of sensor variability using pct error. Error=sensor value - ideal value / ideal value.

Thx, San