Bit error rate
The bit error rate ( BFH ) is a measure of telecommunications for the quality of the transmission over a channel or a digital transmission link
Other names for bit error rate are bit error rate ( BFR ) and bit error rate (BER ) ( with the dimension " number of errors per unit of time "). The unitless BER ( Bit Error Ratio Sheet ) on the other hand, the error quotient from the number of incorrectly received bits in an arbitrary time interval ( bit errors ), and the number in the same time interval of a total received bits is calculated. If it is stored bits on a storage medium, a defined time interval is not based on, but a defined storage size.
For example, a bit error ratio of 3.10 -6 means that over 1 million bits transmitted or stored, an average of 3 bits can be wrong. The measurement for determining the bit error ratio on transmission lines is called BER test or BERT. It is usually carried out with the help of test bit patterns that are transmitted from a measuring device and received by the transmission again and compared.
In many publications is based on the English term bit error rate (BER ) for a bit error rate, the term bit error rate preferred. In many documentaries (including the manuals of Antennenmeßgeräten ) the bit error rate is confused with the bit error ratio. So in digital TV broadcasts " BER " basically means "Bit Error Ratio " and thus the bit error ratio.
In contrast to the bit error rate describes the bit error probability (English bit error probability, BEP) calculated by a theoretical consideration of a bit error probability of the occurrence. Closed calculations of bit error probabilities are usually possible only for idealized scenarios, but are often used to replace expensive simulations and to determine limits of performance of communication engineering systems.
Excessive Bit Error Rate, abbreviated exBER, ( in German: very high bit error rate ) is an error definition, which is used in the synchronous digital hierarchy and the Plesiochronous Digital Hierarchy.
It is the most widely used quantitative feature for the quality of service of the data communication. The exact definition of the error is dependent on technique and bit rate. For example, it is customary to designate 1 bit error in 1000 transmitted bits when the error exBER achieved. The second, in which at least one exBER has occurred, usually recorded and counted in a MIB. With such counters, the quality of a data communication can be measured during the period of use is defined.
A similar definition signal degraded (SD). The error threshold of " signal degraded" is usually achieved at 1 bit error in 1 million bits transmitted. At bit rates of 150 Mbit / s and higher, the error thresholds of exBER and SD are higher by several orders of magnitude. The error exBER is the contractually specified criterion for a failure of the dedicated line for leased lines in Germany.
Measurement methods
The measurement of the bit error rate requires a high-quality reference information. This can for example be
- A comparison of a mitempfangenen check information,
- An idealized model- reference information,
- An evaluation of the received power spectrum,
- An assessment against a known test signal,
- An error-free independent reference information or
- A comparison over a sufficiently long measurement time with repeating information
In each case, a reliable reference value must be used.
Example
Refers to the signal -to-noise ratio on a bit of information, you get the bit energy to noise power density ratio, the ratio of the time spent on an information bit energy Eb to the noise spectral density N0.
With increasing noise, that is, decreasing Eb/N0, the bit error rate increases. Furthermore, the BER is dependent on the coding method. 16 -PSK or 8 -PSK have a higher information density than, for example 2 -PSK and require a higher signal - to-noise ratio. Forward error correction techniques reduce the required signal quality.
The figure lists the error rate as a function of Eb/N0 (shown in dB ) for different performance- limited coding. For example, at a BER of 10-4, the Eb/N0 ratio BPSK about 8 dB with the FEC - encoding decodes convolutional codes with the Viterbi algorithm, about 4 dB. An additional Reed-Solomon coding reduces the minimum required Eb/N0 at less than 3 dB. The vertical line in the figure indicates the Shannon limit that can not be exceeded.