Pulse or digital communications – Testing – Data rate
Reexamination Certificate
1999-12-27
2003-11-25
Phu, Phoung (Department: 2631)
Pulse or digital communications
Testing
Data rate
C375S227000, C714S801000
Reexamination Certificate
active
06654411
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to digital communications systems and, more particularly, to a method and system for estimating the bit error rate (BER) in such systems when the actual BER may have a wide dynamic range.
BACKGROUND OF THE INVENTION
Typically, the communication channel across which a digital signal is transmitted will cause corruption of the signal to some degree. The ratio of corrupted bits to total transmitted bits is known as the bit error rate, abbreviated BER. Generally, it is rare to find a communication channel that has a BER of less than 10
−16
. Thus, a finite bit error rate will affect information travelling on virtually any communication channel.
The public's ever increasing thirst for high-rate data services has resulted in a demand for increased reliability associated with the transport of such services. Thus, it becomes increasingly important for network operators to know the conditions under which their communications networks are operating, so that preventative maintenance or protection switching can be triggered when necessary. Ideally, such actions should be triggered when the BER of the channel exceeds a certain threshold for more than a certain amount of time and it is therefore of paramount importance to obtain a reliable measurement of the BER.
Unfortunately, conventional approaches do not take into consideration the fact that the BER can have an extremely wide dynamic range. That is to say, some conventional methods work well at low error rates and others at high error rates, but typical BER estimation methods do not provide an accurate estimate of the BER across a wide dynamic range.
At low error rates, it is crucial to ensure that each error is captured by the BER measurement system. To this end, most conventional approaches employ parity bits which are part of the overhead of transmitted frames. The receiver then reconstructs the parity operation for each received frame and compares the result to see if it is identical to the parity bits received in the overhead. If not, then an error is declared to have occurred.
In the case of a 40 Gbps signal and Poisson distributed errors having a rate of 10
−16
, the occurrence of two errors takes approximately one week. At such low error rates, errors are most likely to occur very far apart in time, which means that it is extremely rare to have more than one bit error in a group of bits from which a parity bit is computed. Thus, it is possible for the receiver to catch all errors because there is virtually zero probability that the parity bits are unreliable.
However, the just described approach has severe drawbacks at higher error rates. For instance, when the BER increases beyond about 4.6*10
−4
, it can be shown that so many of the bits of an 8000 Hz OC-768 frame are in error that the parity bits in the frame become meaningless. In other words, the parity bits become “saturated” and the receiver is unable to discriminate between a BER of 4.6*10
−4
or any higher BER.
One remedy which prevents the onset of saturation until a BER of about 10
−2
is simply the use of a larger number of parity bits. However, the parity bits might take up more space than is available in the entire overhead of a conventional SONET frame. Furthermore, the receiver must devote such an enormous amount of its resources to processing the parity bits that it will be incapable of performing other tasks such as processing the user information carried by the frame.
Thus, the industry is in need of a technique by virtue of which it would be possible to detect, to within a reasonable degree of accuracy, the BER of a communication channel, where the BER may potentially have an extremely wide dynamic range.
SUMMARY OF THE INVENTION
By virtue of a special parity check performed at the transmitter for each transmitted frame and by virtue of a special processing routine performed at the receiver on each received frame, it is possible to accurately estimate the BER across a wide dynamic range.
According to the invention, a measurement array is created at the transmitter from the bits in each frame, known as the frame bits. The measurement array is constructed from various combinations of frame bits, including at least one combination which includes many frame bits and at least one combination which includes only a few frame bits. The measurement array is then transmitted along with the frame bits, preferably in the overhead of the frame itself. Upon receipt of the frame, the receiver performs a parity check on the received frame bits and on the measurement array, thereby to create a symptomatic array.
The symptomatic array can be processed during the course of successive frames and the results of this processing could be combined and mapped to an estimate of the error rate.
The symptomatic array could also be accumulated during the course of successive frames, resulting in an accumulated symptomatic array. After a sufficiently high number of frames, it will become clear from the values of the elements in the accumulated symptomatic array which bits in the symptomatic array are saturated and which are not.
The accumulated symptomatic array could then be mapped to a BER estimate. This mapping could be based on the proximity of the elements of the accumulated symptomatic array to a set of values associated with a known BER or based on the result of a dot product between the accumulated symptomatic array and a weight vector.
Therefore, the invention may be summarized broadly as a method of estimating the error rate of a communication channel, which includes the steps of
a) for each of a plurality of sub-sets of data elements in the frame, deriving a respective measurement value from the sub-set of data elements, wherein at least two of the sub-sets contain substantially different numbers of data elements;
b) transmitting the data elements and the measurement values across the communication channel;
c) receiving the data elements and the measurement values; and
d) for each received measurement value, deriving a respective symptomatic value from the received measurement value and from the received data elements corresponding to the data elements in the sub-set from which the corresponding measurement value was generated prior to transmission thereof.
The above process is repeated for a number of subsequent frames and the arrays of symptomatic values associated with the various frames are processed to yield an estimate of the error rate.
Preferably, the number of sub-sets associated with each frame is logarithmically related to the number of data elements in the frame. Also preferably, the union of the sub-sets associated with each frame cover all the data elements in the frame.
The step of deriving a measurement value may include determining the parity of the data elements in the respective sub-set of data elements. Similarly, the step of deriving a symptomatic value may include determining the parity of the respective received measurement value together with the received data elements corresponding to the data elements in the respective sub-set from which the corresponding measurement value was generated prior to transmission thereof.
Each frame may include a payload and an overhead and the measurement values associated with a particular frame may be inserted into that frame's overhead. However, the measurement values could be transmitted separately from the data elements.
The step of generating an estimate may include generating a respective intermediate estimate from each symptomatic array and generating the estimate from the intermediate estimates. An intermediate estimate could be generated by comparing the respective symptomatic array to a set of pre-determined vectors associated with respective error rates and choosing as an intermediate error rate estimate the error rate associated with the pre-determined vector being closest to the summed symptomatic array.
Alternatively, the generation of an estimate may include processing the symptomatic arrays in an element
Oberhammer Wolfgang W.
Roberts Kim B.
LandOfFree
Method and system for estimating error rate of a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and system for estimating error rate of a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for estimating error rate of a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3137727