On-line bit error rate estimator

An apparatus for estimating the binary error rate on an input data stream of binary signals whose average maximum voltage, representing a binary "one," is R+V, and whose average minimum voltage, representing a binary "zero," is R-V, comprising:

• a servo loop for establishing values of R-V and R+V, respectively, including a first and second comparator, the input data stream being connected to the inverting terminal of said first comparator and to the non-inverting terminal of said second comparator;
• a first decision flip-flop whose input is connected via a wired OR-gate to the outputs of each of said two comparators;
• a lowpass filter whose input is connected to the output of said first flip-flop;
• an operational amplifier whose non-reference input terminal is connected to the output of said lowpass filter and whose reference terminal is connected via a second lowpass filter to a point which is at voltage R;
• a unity gain inverting amplifier whose input is connected to the output of said operational amplifier;
• a first voltage divider which is connected between the output of said operational amplifier and the non-inverting terminal of said first comparator;
• a second voltage divider which is connected between the reference terminal of said unity gain inverting amplifier and the inverting terminal of said second comparator;
• a power supply whose input is connected to the output of said unity gain inverting amplifier and which removes the value R from the voltage at said terminal;
• a second decision flip-flop, one power supply input terminal of which is connected to the output of said power supply and the data input terminal of which is connected to the input data stream;
• wherein a clock input pulse set at approximately the same rate as the data rate of the input data stream is connected to the clock terminals of each of said first and second flip-flops.

A method for estimating the binary error rate of a binary coded data stream not including code characters of the parity check type, said method comprising:

• establishing from the data stream itself the average maximum voltage deviation with respect to the normal baseline voltage of the input data stream;
• doubling this average deviation to obtain a reference deviation;
• and
• comparing the voltage deviation of the input data stream to this reference deviation at many predetermined sampling points.