Bit Error Rate | previous | next | feedback |
Definition: average fraction of wrongly transmitted bits in a communication link
Even a digital data transmission system is not totally error-free – statistical fluctuations related to noise influences cause a small percentage of the transmitted bits to be corrupted. The average fraction of incorrectly transmitted bits is called the bit error rate. The maximum capacity of a reliable data transmission system is not reached by keeping the bit error rate at an extremely low level (nearly avoiding any bit errors), but by pushing the data rate to a level where some tolerable bit error rate of e.g. 10-12 can be maintained and by detecting and correcting most of the remaining bit errors. For example, detected bit errors may be corrected simply by resending the few affected data packages.
In practice, the bit error rate of a system for optical data transmission (e.g. a fiber-optic link) can be increased by noise influences (particularly in the receiver, but also in the transmitter and in amplifiers), by optical losses, and chromatic and other types of dispersion. Also, nonlinearities can cause signal distortion and channel cross-talk. Possible countermeasures include using a higher transmitted optical power, reducing fiber losses, employing fiber amplifiers and/or dispersion compensation in the link, using improved detectors (possibly with electronic dispersion compensation), optimizing the arrangement of wavelength channels in a wavelength division multiplexing system, or simply by reducing the data transmission rate.
Bit error rates are typically measured with complex devices, called bit error rate testers (BERT), generating a pseudo-random bit sequence and comparing the sent and received data.
See also: optical data transmission, optical fiber communications, electronic dispersion compensation
