BER resolution

[cochenob]

I am using the MT BER VI to measure bit error rate. 

In my current application, I'd like to make BER measurements down to 10^-6 or 10^-9, however it seems that the lowest resolution possible is only around 10^-4 or so.

So...I though to get any information between 10^-6 to 10^-9....I need to increase the number of bits examined at a time.

This however is a challenge, as most of the MT VI's start to choke when presented with more than a few thousand bits (the PN Generators, the demodulation VI's, the measurement VI's, etc..).  If you only present 10000 bits, you'll only get 1/10000=10^-4 resolution.

I considered sending many "packets" of bits, to try and let things add up.  However, the BER at any one time will be dependent on the number of bits it is presently looking at.  As such...the accumulated BER won't be of any higher resolution.

Is there any way around this?

---

Brandon

[mikeporter]

You haven't said what sort of communications link you are concerned with, but the thing to remember is that BER testing is essentially a statistical process. Just because you pass 10000 bits and saw 1 bit error, that doesn't mean that this error level is typical for the communications link. The more data you pass through a link the greater will be your certainty that the measured error rate is typical. As long as the conditions remain unchanged over the communications link it doesn't really matter how long it takes to pass the data, or how you divide it up into blocks or packets.

Mike...