Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
2001-04-17
2002-07-02
Tu, Christine T. (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
C714S701000
Reexamination Certificate
active
06415412
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a signal used in the transmission of data in a communication system, wherein the signal is comprised of continuous, successive cells. The invention further relates to a method for synchronizing a receiver relative to the cells of such a signal, a method for generating the above-named signal, and transmitting and receiving circuits used to implement this method.
BACKGROUND OF THE INVENTION
In packet-oriented data transmission, such as, for example, in an ATM system (ATM=asynchronous transfer mode), the receiver must recognize (in English, packet delineation) the start time of the incoming packets (also called cells). The quality of this synchronization significantly affects the overall performance of the system. A poor “packet delineation” increases proportionately the number of data packets that are lost, and can thereby, for example, destroy the performance of other processing steps.
Cell synchronization should—first of all—be as immune as possible to bit errors and bit patterns that may be present within the transmitted data (so-called mimics: random patterns that simulate a synchronization pattern) and will lead to synchronization error. Second, in the case of a synchronization error, the receiver should be capable of extricating itself from this situation independently. Third, and finally, it should be possible to implement the method at the lowest possible cost in terms of hardware. Generally, the cost in terms of hardware increases proportionally to reliability requirements. To this extent, then, the first two requirements stand in direct contrast to the third requirement.
In communication systems (especially in networks having several nodes), in addition to the user data, monitoring and control data usually must also be transmitted. The monitoring data may be based upon the user data or some given system conditions.
The system user is interested primarily in having his data transmitted at the highest possible speed. It would thus be to his benefit to have the largest possible band width at his disposal. For physical reasons, or due to standardized requirements, however, the overall band width that is available is limited. The transmission of control data can thus result in a necessary reduction in the band width available for user data. In order to leave the largest possible band width available to the user, the monitoring and/or control data should be transmitted as efficiently as possible.
SUMMARY
It is the object of the invention to provide a signal of the type described above such that a high probability of synchronization is ensured and synchronization errors can be detected within a short period of time. In addition, it is the object to process the signal at little cost in terms of circuit engineering.
In accordance with one embodiment of the invention, at least some of the transmitted packets or cells contain a predefined bit pattern, which is recognized by the receiver and which permits the detection of the cell boundaries. In order to prevent the receiver from mistakenly synchronizing itself to a bit pattern that may be contained in the (randomly variable) data stream from the user, the communication system, at the transmitter end, provides the synchronization bit pattern with a specific error protection code. This error protection code differs from the code that is used to protect the user data.
Recognition of a predefined bit pattern is relatively simple and quick. This is important for communication systems having high data transmission rates. If the sequence of the data bits to be transmitted by the user should coincidentally be identical to the predefined synchronization pattern (mimic), this will necessarily result in a different error protection code. If the synchronization pattern and its error protection code are regarded as a single unit, then the invention will ensure (with very high probability) that this unit cannot be simulated by the user.
The predefined bit pattern should be of a certain length, so that the probability that precisely this bit sequence will occur in the data from the user is very slight. It will typically have a length of several bytes. Due to the encoding specified in the invention, it is not necessary for the length to be greater than 20 bytes. In the exemplary embodiment described in detail below, for example, this length is 10 bytes.
Preferably, both the predefined bit pattern and the user data are provided with a parity code. In accordance with the invention, however, various polarity patterns are used. If, for example, the polarity pattern of the user data alternates even and odd, then that bit pattern may have several equal polarities one after another.
Of course it is also possible for the predefined bit pattern on the one hand and the user data on the other hand to be processed with totally different encoding procedures. What is important here is that the probability is infinitesimal, if not zero, that a coincidentally identical bit pattern in the user data would result in the same error protection code.
It naturally remains a possibility that a mimic may be produced as a result of a certain error combination in transmission. In order for the receiver to be able to recognize an error in synchronization, the transmitted cells are also provided with a block code, at a predetermined position, which essentially encodes the payload region of the cell. In the case of a synchronization error, the receiver naturally interprets any byte as a block code, so that, as a result, the bit error rate will lie far above the average bit error rate for the system.
A particularly simple and effective encoding is presented by bit interleaved parity encoding. In this, in principle, all bytes of the cell are encoded in columns. The BIP value (BIP=bit interleaved parity) is positioned, for example, at the end of the cells. It can be advantageous for the BIP value not to encode all of the bytes of the cell. According to a preferred embodiment of the invention, for example, one report byte is not encoded in the BIP value. This byte can then be changed from cell to cell in order, for example, to permit recognition of a so-called “slip.” (In a “slip” the receiver misses individual time cycles for one reason or another. The report byte makes it possible to recognize a “slip” even when the user data, for example, contain a series of zeros).
In accordance with one preferred embodiment of the invention, the predefined bit pattern is transmitted not in those cells that contain the user data, but rather in separate filler cells. These are always transmitted when no user data are present, but, also when the capacity of the transmission link has been exceeded. In each case, these filler cells should be present in the data stream at a frequency rate of at least 0.1%. If—viewed statistically—fewer filler cells are present, then the synchronization time will be disproportionately high (because the synchronization bit pattern will appear too infrequently). Advantageously, the statistical frequency lies within a range of 1%.
Naturally, the synchronization bit pattern should not be transmitted more frequently than necessary. If it were present in more than 10% of the cells, then the effects on transmission capacity would be noticeable.
It is advantageous for the length of the cells to be firmly predetermined (for example 53 bytes). It is, however, entirely possible to work with cells of varying lengths. Obviously, some indication of the length of the given cell must then be provided somewhere in the cell header.
In the simplest case, filler cells and data lines are equal in length and of basically the same format (for example 5 bytes header, 48 bytes payload). The predefined bit pattern is then positioned in the header of the filler cells, and the BIP value is at the end of each data line. In principle it remains possible for the synchronization bit pattern to be contained in the header of a data cell. With relatively small cells (such as are typically found in ATM systems), however, this would take
Ascom Tech AG
Birch & Stewart Kolasch & Birch, LLP
Tu Christine T.
LandOfFree
Signal used in the transmission of data in a communication... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Signal used in the transmission of data in a communication..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Signal used in the transmission of data in a communication... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2903680