Chuck:Quote:
Originally Posted by
Calaveras
Actually that's not clear to me at all. When I was performing these measurements in the lab we'd just give the number as a kind of shorthand. The "/sec" part was always left out so I can't tell from above what they really mean. I just assumed what was meant was 3x10E-6/sec.
I'm probably being too picky, and it doesn't make much difference when everyone uses the same measurement expressions for comparisons, but let me try again to make the distinction between Bit Error Ratio and Bit Error Rate by using a car analogy:
RATIO: So far, I have driven 5 miles of the 100 miles to Richmond.
Notice there is no mention of how fast I'm driving, just as there is no mention of how fast the errors are happening in Bit Error Ratio.
RATE: I can drive 60 miles per hour when I travel to Richmond.
Notice that there is no mention of how many miles to Richmond, just as there is no mention of total bits in Bit Error Rate.
If I may quote authorities who are a lot smarter than I am:
Digital Communications Test and Measurement: High-Speed Physical Layer Characterization
by Dennis Derickson and Marcus MÃ¼ller
Publication Date: December 20, 2007 | ISBN-10: 0132209101 | ISBN-13: 978-0132209106 | Edition: 1
Dennis Derickson is an assistant professor at California Polytechnic State University. He spent eighteen years as member of technical staff and project manager at Hewlett-Packard and Agilent Technologies before serving as applications engineering manager for Cierra Photonics. He has authored or coauthored fifty publications in high-speed communications and is the editor of Fiber Optic Test and Measurement (Prentice Hall, 1998). Dennis has a Ph.D. from the University of California, Santa Barbara.
Marcus MÃ¼ller is an R&D lead engineer with Agilent Technologies' High-Speed Digital Test segment in Boeblingen, Germany. He specializes in bit error ratio and jitter analysis of high-speed links, and has contributed to new methods for total jitter measurement at low bit error ratios, and jitter tolerance test. Marcus received his M.Sc. degree from Stuttgart University, Germany, in 1999.
Chapter 4 (bit error ratio testing): A fundamental property of a digital communications link is the bit error ratio (BER), which is the number of bit errors divided by the total number of bits sent. This chapter along with Appendix A provides a detailed description of the test hardware and test methodology for BER. (Marcus MÃ¼ller wrote this chapter.)
Quote:
The bit error ratio (BER) is a measure of the percentage of bits that a system does not transmit or receive correctly. It is a dimensionless number, ranging from 0.0 to 1.0: If the BER = 0.0, then all bits are transmited correctly; at the other extreme, if the BER = 1.0, every bit is received in error.
The bit error ratio is calculated by dividing the number of erroneous bits by the number of compared bits.
A second measure for the error performance of a digital transmission system is the bit error rate. It's different from the bit error ratio in that it relates the number of errors to the test time, rather than the number of bits compared during the test. Note that bit error rate is not dimensionless; its unit is errors per time.
In everyday use bit error rate is more descriptive than the bit error ratio: A PCI Express link, for example has a target ratio of 10E-12. At the 2.5 Gbit/s data rate, that's an error rate of 0.025 errors per second.
The terms bit error rate and bit error ratio are often mixed up in everyday usage, especially since the BER acronym is used for both. However, since the bit error rate is hardly used in practice these days, it is usually safe to assume that bit error ratio was intended.
In a earlier post I said that ATSC A74 used a number for Bit Error Rate that was really Bit Error Ratio:
Quote:
A DTV receiver should achieve a bit error rate in the transport stream of no worse than 3x10E-6 (i.e., the FCC Advisory Committee on Advanced Television Service, ACATS, Threshold of Visibility, TOV) for input RF signal levels directly to the tuner from â€“83 dBm to â€“5 dBm for both the VHF and UHF bands.
As Marcus MÃ¼ller said, Bit Error Ratios are dimensionless, which means that there are no units, so you shouldn't add per sec. The reason Bit Error Ratio is dimensionless is because when you divide error bits by total bits for the ratio, the bits units cancel according to math rules, and you are left with a small number divided by a large number which is conveniently expressed in scientific notation like, in this case, 3x10E-6.
In an earlier
post dr1394 conveniently showed us:
# of Pre RS errors/sec
<1
<10
<1000
<10000
<100000 heavily impaired
That's Bit Error Rate.
Q.E.D. (This QED is not a ham Q signal)
73,
rabbit
W4...