Bit error rate detection

Details Bit error rate detection Bit error rate detection

1999-11-19

2003-07-08

Dildine, R. Stephen (Department: 2133)

Error detection/correction and fault detection/recovery

Pulse or data error handling

Error count or rate

C714S802000

Reexamination Certificate

active

06591383

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates generally to high-speed digital communications, and specifically to SONET/SDH communication networks.
BACKGROUND OF THE INVENTION
SONET (Synchronous Optical Network) and SDH (Synchronous Digital Hierarchy) are two families of closely-related and mutually-compatible standards for high-speed data communication. These standards govern interface parameters; rates, formats and multiplexing methods; and operations, administration, maintenance and provisioning for high-speed signal transmission. SONET and SDH use frame formats for transmitting data. SDH/SONET requires that errors in transmission be detected and an alarm condition be set when a bit error rate in a data stream exceeds a specified level. The bit error rate (BER) is defined as the number of erred bits in a frame divided by the total number of bits in the frame. Typically, the BER alarm is an indication that a fiber channel over which the transmission is being conveyed has degraded, and the transmission is then switched to a different, undegraded channel.
The maximum permitted BER is a function of line type. Typically a defect condition is detected when the BER exceeds 10
−3
. A hysteresis is introduced in clearing the defect condition, so that typically only when the BER drops back below 10
−4
is the defect condition cleared. According to requirements set by the International Telecommunications Union (ITU) in standard G.783, detection and clearing of the defect state must take place within a specified detection time. Other requirement levels have different BER limits, such as 10
−4
or 10
−5
, and different detection times.
SDH/SONET data frames include bytes set aside for error detection using bit interleaved parity (BIP) checks. For each frame, the BIP is calculated before transmission and is then transmitted with the next frame. At the receiving end, the BIP is recalculated and compared with the transmitted BIP value in the next frame. A discrepancy between recalculated and transmitted BIPs is noted as a BIP error. It is intuitively clear that the number of BIP errors in a given time period is a function of the current BER, and therefore BIP error detection is used for the purpose of BER defect detection and clearing. The relation between BIP errors and BER is probabilistic, so that a certain amount of uncertainty is inevitable in using BIP errors to track BER defects. The ITU requires that when BER≧10
−3
, BIP processing (or any other estimation method that is used) detect the defect condition within the permitted detection time (40 ms in one specific case) with a detection probability of ≧0.99, whereas when BER≦10
−4
, the false alarm probability, i.e., the chance of incorrectly reporting a detected defect condition, should be ≦10
−6
. By the same token, when BER≦10
−4
, the probability of clearing an existing defect condition should be ≧0.99, whereas the probability of clearing while BER≧10
−3
should be ≦10
−6
. The required calculations must be performed at high speed and impose a heavy load on the hardware and software responsible for BER defect detection.
A number of methods have been proposed to reduce the hardware and software resources required for BER detection. For example, U.S. Pat. No. 5,764,651, whose disclosure is incorporated herein by reference, describes a system for finding an error rate in a data stream using a variable length window to sample erred bits in the data stream. If the measured error rate, based on BIP calculation, is greater than an allowable threshold, the window length is reduced. When the window length reaches zero, a defect condition is detected.
European Patent Application EP 862,292, whose disclosure is incorporated herein by reference, describes a method for BER detection using two or more BIP monitors with respective fixed window lengths. Each of the monitors detects the presence or absence of a particular BER, for example, 10
−6
and 10
−7
, respectively, by integrating BIP errors over time. A monitor client reacts to notifications from the BER monitors so as to determine that the BER state of the data stream has changed and to generate an alarm when appropriate.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide improved methods and apparatus for monitoring the BER of a data stream.
In some aspects of the present invention, the methods and apparatus are applied to a SDH/SONET data stream so as to detect and clear BER defect states in accordance with accepted communications standards.
It is a further object of some aspects of the present invention to provide reliable and flexible methods and devices for detecting and clearing BER defect states based on BIP error detection.
In preferred embodiments of the present invention, an error detection method for detection of BER conditions in a SDH/SONET data stream comprises first and second processing stages. In the first stage, data blocks of given size are successively processed to determine a measure of bit errors in each block, preferably based on the number of BIP errors in the block. If the error measure in a given block exceeds a first predetermined threshold, the block is noted as an error block. In the second stage, the error results are stored, preferably in a shift register, for an interval of time containing a selected number of successive blocks. The shift register thus defines a “rolling window” in time. If the number of error blocks within the interval exceeds a second threshold, then a BER defect condition is detected. The size of the data blocks, the duration of the window, and the first and second thresholds are chosen based on the statistical relationship of the number of BIP errors to the BER of the data stream, so that the actual BER condition is detected within the time permitted and with the levels of confidence (i.e., high detection rate and low false alarm rate) dictated by applicable standards.
Once the BER defect condition is detected, one or more of the detection parameters, preferably including the second threshold, are readjusted. The new second threshold represents a maximum number of error blocks in the rolling window for which the error condition will be maintained. When the number of error blocks in the window drops below this threshold, the BER condition is cleared. Optionally, other parameters, such as the block size, window length and first threshold, may also be changed when seeking to clear the BER defect condition. Maintaining these other parameters at the same values for both the defect detection and defect clearing phases of operation is generally advantageous, however, since it simplifies the hardware and software design of the device.
Preferably, for the sake of computational simplicity and speed, the errors are determined with respect to a BIP
2
calculation, comprising a 2-bit result in which one bit is the XOR of all of the even bits in a given data frame, and the other bit is the XOR of all of the odd bits. Alternatively, however, any other type of BIP calculation, such as BIP8 (yielding an eight-bit result) or error detection calculations of other types, may be used.
There is therefore provided, in accordance with a preferred embodiment of the present invention, a method for detecting an error rate of a data stream, including:
dividing the data stream into a sequence of blocks;
defining a detection interval including a predetermined number of blocks in the sequence;
computing for one or more of the blocks in the detection interval respective error measures responsive to the error rate of the data stream;
classifying the one or more blocks in the detection interval as good or bad blocks by comparing the respective error measures to a first threshold; and
estimating that an error condition exists in the data stream by comparing a count of the bad blocks in the interval to a second threshold.
Preferably, computing the error measures includes computing bit interleave parity (BIP) values and comparing the computed BI

Affiliated with

Also associated with

Rating

Bit error rate detection does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Bit error rate detection, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Bit error rate detection will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3005817