Error-robust multiplexing method
Abstract
The method involves transmitting data in data blocks between two stations, whereby the data blocks comprise a synchronization pattern (SYNC), a header (HEADER), and a data field (INFORMATION). The synchronization pattern indicates the beginning of the data block, and the header field includes control symbols for the use of the following data field. At least one synchronization pattern and a header free of errors have to be received for an initial synchronization.

Term
Term ended
Projected expiry passed 20 March 2017, 9.5 years ago.
- Priority
- Filed
- Published
- Projected expiry
- Today
19 claims: 15 independent, 4 dependent
- 1Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen, wobei die Datenblöcke ein Synchronisationsmuster SYNC, einen HEADER und ein INFORMATIONS-Feld aufweisen, wobei das Synchronisationsmuster SYNC den Beginn des Datenblockes anzeigt und der HEADER Steuerzeichen für die Behandlung des nachfolgenden INFORMATIONS-Feldes enthält, dadurch gekennzeichnet, daß für die Startsynchronisation mindestens das Synchronisationsmuster SYNC und ein fehlerfreie HEADER gefunden werden muß.
- 2Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach Anspruch 1 dadurch gekennzeichnet, daß als Startsynchronisation für mehrere aufeinanderfolgende Datenblöcke das Synchronisationsmuster SYNC und der fehlerfreie HEADER des letzten Datenblocks gefunden werden müssen
- 3Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach Anspruch 1, dadurch gekennzeichnet, daß als Startsynchronisation zu einem Datenblock das Synchronisationsmuster SYNC und der dazugehörige fehlerfreie HEADER gefunden werden muß.
- 4Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach Anspruch 1 und 2, dadurch gekennzeichnet, daß die Startsynchronisation G1-mal wiederholt wird.
- 5Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 1 und 2, dadurch gekennzeichnet, daß die Synchronisation als verloren gilt, wenn SYNC und der dazugehörige HEADER in einem Datenblock nicht gefunden werden können, oder wenn mehrfach bis zu einem Grenzwert G2 die Synchronisation nicht gelingt.
- 6Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 1 bis 5, dadurch gekennzeichnet, daß das Synchronisationsmuster SYNC über Korrelationsfunktionen erkannt werden muß.
- 7Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 1 bis 6, dadurch gekennzeichnet, daß zur Erkennung von Fehlern im HEADER Fehlererkennungsverfahren angewendet werden.
- 8Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen, wobei die Datenblöcke ein Synchronisationsmuster SYNC, einen HEADER und einen INFORMATIONS-Feld aufweisen, wobei das Synchronisationsmuster SYNC den Beginn des Datenblockes anzeigt und der HEADER Steuerzeichen für die Behandlung des nachfolgenden INFORMATIONS-Feldes enthält, dadurch gekennzeichnet, daß das INFORMATIONS-Feld mit Füllinformationen aufgefüllt wird, wenn die sendende Datenquelle die eingestellte Länge des Datenblocks nicht mit Daten füllen kann.
- 9Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach Anspruch 8, dadurch gekennzeichnet, daß nach einer vollständigen Erzeugung einer Rahmenstruktur der Datenquelle, der restliche Multiplexslot mit Füllinformation aufgefüllt wird, wenn die Datenquelle keine Daten mehr sendet.
- 10Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach Anspruch 8, dadurch gekennzeichnet, daß nach einer nicht vollständigen Übertragung einer Rahmenstruktur der Datenquelle die Informationen zwischengespeichert wird und der Multiplexslot mit Füllinformation aufgefüllt wird oder ein anderer HEADER gewählt wird.
- 11Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 8 bis 10, dadurch gekennzeichnet, daß die Füllinformationen aus Synchronisationsworten der Datenquelle bestehen.
- 12Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 9 bis 11, dadurch gekennzeichnet, daß das Einfüllen von Füllinformationen im HEADER signalisiert wird.
- 13Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 9 bis 12, dadurch gekennzeichnet, daß das Einfüllen von Füllinformationen im HEADER mit t Bits signalisiert wird.
- 14Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 9 bis 13, dadurch gekennzeichnet, daß mit t Bits Datenquellen spezifiziert werden, für die die Füllinformationen benutzt werden.
- 15Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 8 bis 14, dadurch gekennzeichnet, daß die Füllinformationen aus mehreren aufeinanderfolgenden Bits besteht, vorzugsweise aus alles Null"(0000 0000) bzw. alles Eins" (1111 1111).
- 16Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 9 bis 15, dadurch gekennzeichnet, daß das Einfüllen von Füllinformationen im Zusatzheader der einzelen Quelle angezeigt wird.
- 17Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 1 bis 16, dadurch gekennzeichnet, daß im Demultiplexer Softbits" weiterverarbeitet werden.
- 18Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 1 bis 17, dadurch gekennzeichnet, daß Fehlererkennungsverfahren und eventuell Fehlerkorrekturverfahren auf den gesamten Datenblock angewendet werden.
- 19Verfahren zur Datenübertragung mittels Datenblöcken zwischen zwei Stationen nach den Ansprüchen 1 bis 18, dadurch gekennzeichnet, daß die Informationen des HEADERs über den gesamte Datenblock verteilt werden (Interleaving).
Independent claims19
40 paragraphs, as filed
State of the art
0001The invention is based on a method for the transmission of data blocks according to the preamble of the main claim.
0002A method for transmitting data blocks is already known, the protocol H.22P (ITU-T Study Group 15, LBC 95-276<img file="EP0802650A2_D0001.tif" />ITU-T Recommendation H.22P ") In this protocol, the frame structures, the formats of the data and control fields and a structure for the data to be transmitted by the multiplexer, the multiplex protocol, are specified. The multiplex protocol enables the processing of logical information that reaches the multiplex level via the adaptation level into uniform data units. The protocol enables the transmission of any combination of digital audio and video data or other information via a data line and, to prevent data loss, proposes a special protocol that has a synchronization pattern with a length of 31 bits. It is followed by the HEADER (31 or 63 bits) and the information field with a fixed length. The synchronization pattern must be recognized by a correlation condition in the receiver, only then can the processing of the data blocks in the receiver begin. With this method, data loss can occur due to a loss of synchronization. In addition, there is no mechanism that fills fixed-length data blocks with data if the data source stops sending data.
Advantages of the invention
0003The inventive method with the characterizing features of claim 1 has the advantage that the synchronization is combined with an error detection method that detects errors in the transmission of the HEADER. This means that an error-free HEADER can be used to ensure synchronization.
0004The measures listed in the subclaims make particularly advantageous developments and improvements of the method specified in claim 1 possible.
0005The start synchronization process, which shows a very fast possibility of synchronization, is particularly advantageous.
0006For safety reasons, the start synchronization is advantageously carried out a number of times.
0007In the case of multiple synchronization, it is also sensible to define a limit value according to claim 5.
0008It is important for the security of the method that the SYNC is recognized by a correlation function.
0009The fast establishment of the connection and a successful start synchronization are only guaranteed by the error detection process.
0010It is advantageous for problem-free data transmission to replace missing information from the data source with filler information.
0011It is advantageous to fill the INFORMATION field with filling information if the data source has still sent a complete frame.
0012If this is not the case, the data will be cached.
0013A very good way of filling up the data is to feed synchronization words of the data source through the multiplexer.
0014The signaling that fill bits have been used has a very favorable effect in the HEADER.
0015Data security can be further improved by using soft bits.
0016The embodiment which extends error detection and error correction methods to the entire data block is also advantageous for secure transmission.
0017Another good way to get an error-free method is to distribute the HEADER over the entire data block (interleaving).
drawing
0018An embodiment of the invention is shown in the drawing and is explained in more detail in the following description. Show it:<dl id="dl0001" compact="compact"><dt>Fig 1:</dt><dd>Structure of the data levels for multiplex data transmission</dd><dt>Fig. 2:</dt><dd>Structure of the data unit of the multiplex protocol</dd><dt>Fig. 3:</dt><dd>Framework structure of the data sources</dd></dl>
Description of the embodiment
0019The data transmission of any data signals takes place according to Fig. 1 over hierarchically structured levels. The signals, some of which are analog, come from the individual data devices via the application level to the coding level. After a digitization step, the contents of the logical channels LCN are passed on to the adaptation level of the multiplexer. The data are still forwarded to the multiplex level as MUX-SDUs (Service Data Units) in separate channels. This level processes the multitude of channels into one channel and outputs MUX PDUs (Protocol Data Units). These data blocks are filled with signals according to the protocol according to the invention.
00202 shows such a MUX PDU in the sequence of the control and data fields. The first field SYNC contains a synchronization pattern of variable length, which contains a clearly detectable bit sequence of, for example, 31 bits. The synchronization pattern is placed in front of each multiplex packet. As a pattern, e.g. B. Barker or Williard sequences can be used. In the HEADER field you define the transmission scheme for the following information block. An example of such a transmission scheme is described in protocol H.223 (ITU-T Study Group 15). Such a HEADER has 4 bits, for example. All 16 states that the HEADER can describe with the 4 bits are stored in a table. For example, if you only want to transmit audio signals, a certain bit sequence is set, if the information block for audio and video signals is shared, another bit sequence is sent. The information field follows. It is structured according to the rules defined in the HEADER for the various data sources. The information field is filled with data in accordance with the multiplex scheme specified by the HEADER until the packet length n is reached. A connection must be established as the first step for data transmission. For this purpose, the length n of the data blocks is determined using a control protocol. The length n can be set for the receiver and transmitter, even at a later time. To do this, the control protocol must intervene in the transmission and carry out a comparison.
0021Since the frame length n of the data blocks is kept constant over a period specified in the control protocol, a particularly simple and fail-safe synchronization can be achieved by always looking for a synchronization pattern (SYNC) at the end of a data block. However, if a synchronization pattern is identified incorrectly after the connection has been established, this synchronization strategy means that a subsequent synchronization pattern cannot be recognized, since it is not in the correct time window. The synchronization of the first data block, the start synchronization, is therefore of particular importance and it must be ensured that this start synchronization takes place with particular certainty. For the start synchronization, it is suggested to include the HEADER in the synchronization, ie a start synchronization can only take place if both synchronization patterns and an error-free HEADER are found. A certain number of incorrect synchronization patterns or HEADERS can then be accepted for the synchronization of the subsequent data blocks (MUX-PDUs) without the synchronization being lost as a result. It is when establishing a connection and the start synchronization z. B. advantageous to use shorter lengths for n at the beginning of the transmission (connection establishment with the help of a control protocol). With the help of the control protocol, a point in time is then defined from which the length n is changed.
0022To identify the SYNC on the receiver side of the data transmission in the demultiplexer, a minimum number of bits is specified which must match between a pattern in the data stream and the synchronization pattern defined by the control protocol. If this correlation minimum is reached (correlation condition), the synchronization pattern is considered to have been found. If a synchronization pattern has been found in this way, an error detection procedure for the HEADER is carried out. The synchronization in the receiver is only successfully established if an error-free HEADER has been found for the synchronization pattern. In the simplest case, the error detection can be a parity check, but is preferably carried out using a CRC code. If an error was found in the transmission of the data, the synchronization process continues with the search for the next synchronization pattern. In this example, the SYNC and the associated error-free HEADER must be found once in order to initiate the start synchronization.
0023If the start synchronization process has been carried out successfully, the next synchronization pattern is searched for a complete frame of length n. At the same time, a counter runs, which is incremented if the synchronization pattern does not meet the correlation condition and the HEADER cannot be recognized without errors. If the counter has exceeded a limit value G2 (integer value that is determined by the control protocol), the synchronization is considered lost and it must be re-synchronized according to the above scheme. After four attempts, the synchronization is typically considered lost and a new start synchronization takes place. If the synchronization in the receiver has been successful, the processing of the HEADERS and the INFORMATION begins.
Further examples:
0024It is advantageous to make the start synchronization safer against errors. The synchronization principle can therefore be expanded by not only finding the synchronization pattern and the associated error-free HEADER once, but G1 times (whole number to be set) before the synchronization is considered successful.
0025A further advantageous embodiment varies the start synchronization, in that the synchronization pattern SYNC must be found for successive multiplex packets and a free HEADER for the last packet.
0026A further embodiment is used in particular when there is not enough information in the data source to fill the available space in the INFORMATION field. When using constant multiplex lengths n, the amount of data to be filled into a multiplex packet, the so-called multiplex slot, is known for each data source. If a data source stops its data transmission, it makes sense to change the HEADER and therefore no longer provide space for the data source in the next information field. If residual data from the data source is to be transferred, it is possible to use suitable filler bits and fill the multiplex slot. To do this, the framework structure of the data streams is used in a suitable manner. The frame structure of the data streams is e.g. B. at the adaptation level of the data transmission, see Fig. 1. In the special case of the data protocol H.324, which specifies the frame structure of the data sources, the procedure can be as follows: The frame structure of the data is limited, for example, by synchronization words of the data source, see FIG. 3. How the frame structure looks in detail is described in protocol H. 245, H.263, G.723, as well as in data protocols. At least one synchronization pattern is set on a frame boundary. The control data contain, for example, information about the type of data source or contain error protection. The information block contains the data that are actually to be transmitted. The frame structure is fed into the multiplex packet, FIG. 1, in the INFORMATION field, ie there are several frames within the INFORMATION field can be located.
0027If a data source has supplied too little data to completely fill the associated multiplex slot in the INFORMATION field, the following strategies are possible in order to still be able to create a complete multiplex package:<ul id="ul0001" list-style="none" compact="compact"><li>a) In the first case, it is assumed that the data source feeds data up to a frame end, but that the INFORMATION field of the data block is not filled with data. The data source sends data synchronization word and possibly other data. The multiplexer fills this data synchronization word into the multiplex slot. After the last transmitted frame boundary, the multiplex slot is filled with the special synchronization word of the data source. Filling with synchronization words takes place in the multiplexer itself. The rest of a synchronization word that has not been completely filled in should be filled in in the next multiplex slot of the corresponding data source.</li><li>b) If the case occurs that the synchronization word for the identification of the frame boundary is missing from the data source, one has to proceed differently. In this case, the existing part of the data must be buffered until the data source has supplied the rest of the frame or a complete multiplex slot to the multiplexer. It is not possible to insert a synchronization word in the information field in FIG. 3 if the data stream is interrupted, since the recipient cannot recognize this word correctly at this point. After the storage process, the multiplex slot is filled with synchronization patterns of the associated data source by the multiplexer or another HEADER is selected that does not contain the relevant data source.</li></ul>
0028If additional synchronization words of the data source have been inserted according to the above procedure, it makes sense to signal them with an additional bit HF in the HEADER.
0029With the help of the HF bit inserted in the HEADER, a further filling mechanism can be carried out. For each data source that has reserved a multiplex slot in the data block according to the multiplex scheme, which it cannot fill completely, the following scheme is used:<ul id="ul0002" list-style="none" compact="compact"><li>a) the first byte of the multiplex slot <img file="EP0802650A2_D0001.tif" />all zero "(0000 0000) is set when the rest of the multiplex slot can be completely filled with data.</li><li>b) If more than 1 byte has to be filled in the multiplex slot, the corresponding bytes are added <img file="EP0802650A2_D0001.tif" />all one "(1111 1111). The last byte before the start of the data is always open <img file="EP0802650A2_D0001.tif" />all zero "(according to a), e.g.</li></ul><tables id="tabl0001" num="0001"><img file="EP0802650A2_D0002.tif" /></tables>
0030If the receiver uses HF to recognize that filling information is in the data packet, this can be removed taking into account the filling rules according to a) and b). However, if more than 3 transmission errors occur per fill byte, it is no longer possible to reliably recognize the fill information. The markings with <img file="EP0802650A2_D0001.tif" />all zero "and <img file="EP0802650A2_D0001.tif" />all one "can also be reversed in their meaning. It is conceivable to evaluate only 7 of the 8 bits of the filler byte, since the eighth bit is redundant anyway in the event of transmission errors.
0031Another possibility is to replace the individual bit HF with t bits. For example, t bits can be used to specify the data source for which filler bits were used.
0032In the following example, HF consists of 2 bits. The following coding is possible with these:<tables id="tabl0002" num="0002"><table frame="all"><tgroup cols="2" colsep="1" rowsep="0"><colspec colnum="1" colname="col1" colwidth="78.75mm" /><colspec colnum="2" colname="col2" colwidth="78.75mm" /><tbody valign="top"><row><entry namest="col1" nameend="col1" align="right">00</entry><entry namest="col2" nameend="col2" align="left">no fill information in the data packet</entry></row><row><entry namest="col1" nameend="col1" align="right">01</entry><entry namest="col2" nameend="col2" align="left">Fill information for audio or video source</entry></row><row><entry namest="col1" nameend="col1" align="right">10</entry><entry namest="col2" nameend="col2" align="left">Fill information for data and control</entry></row><row rowsep="1"><entry namest="col1" nameend="col1" align="right">11</entry><entry namest="col2" nameend="col2" align="left">Fill information for all sources</entry></row></tbody></tgroup></table></tables>
0033The filling method with the zero and inserted bytes can also be used if no bit is set in the HEADER of the data block that bears information about filling bits. It is then conceivable that in one <img file="EP0802650A2_D0001.tif" />Additional header "of the data source contains the information about the filler bytes. This means that the process of filling with filler bits would be independent of the HEADER of the multiplexer. The individual data sources then have information about a possible filler with filler bits via their own additional header. It is also possible that the Additional headers can be defined by the multiplex protocol in the data slots of the data sources.
0034To further improve the synchronization in the multiplexer, it may be useful to further process the so-called soft bits of the demodulator in the multiplexer. This means that symbols are passed on instead of bits. If the demultiplexer forwards the soft bits to the data source, improved FEC methods can be used (soft decision signal processing). The symbol format closes the special case of<img file="EP0802650A2_D0001.tif" />Hard decesion "if only the MSB of a symbol is used again.
0035In order to make the multiplex scheme as secure as possible, it should be transmitted together with an error-detecting code (e.g. CRC) and then an error-correcting code (e.g. BCH, Reed-Solomon, RCPC). In addition, it makes sense to arrange the HEADER over the multiplex package (interleaving).
14 sheets
Sheet 1 Sheet 2 Sheet 3 Sheet 4 Sheet 5 Sheet 6 Sheet 7 Sheet 8 Sheet 9 Sheet 10 Sheet 11 Sheet 12 Sheet 13 Sheet 14
Every citation, both ways
| Document | Relation | Office | Cited during |
|---|---|---|---|
| EP0371593A2 | Cites | European Patent Office (EPO) | Search report |
| US4905234A | Cites | United States of America | Search report |
| US5426643A | Cites | United States of America | Search report |
8 priority claims, no other members on record
Priority claims8
| Document | Office | Kind | Date |
|---|---|---|---|
| 19614738 | Germany | A | |
| 19614738 | Germany | – | |
| 19618386 | Germany | A | |
| 19618386 | Germany | – | |
| DE1996114738 | – | – | – |
| DE1996118386 | – | – | – |
| 19614738 | – | – | – |
| 19618386 | – | – | – |
8 legal events, as the office reported them to INPADOC
Over the term
Point at a mark for the eventEvents
| Event | Code | |
|---|---|---|
| Application deemed to be withdrawnWithdrawn18D | 18D | |
| Information on the status of an ep patent application or granted ep patentGrantedSTATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWNSTAA | STAA | |
| Request for examination filed17P | 17P | |
| Information provided on ipc code assigned before grant7H 04L 1/00 A, 7H 04L 7/04 B, 7H 04L 1/08 B, 7H 04L 7/10 BRIC1 | RIC1 | |
| Designated contracting statesAK | AK | |
| Search report despatchedORIGINAL CODE: 0009013PUAL | PUAL | |
| Designated contracting statesAK | AK | |
| Public reference made under article 153(3) epc to a published international application that has entered the european phaseORIGINAL CODE: 0009012PUAI | PUAI |
Numbers
- Publication
- 0802650
- Publication, DOCDB
- 0802650
- Publication, EPODOC
- EP0802650
- Application
- 97104733
- Application, DOCDB
- 97104733
- Application, EPODOC
- EP19970104733
Titles3
- German
- Fehlerrobustes Multiplexverfahren
- English
- Error-robust multiplexing method
- French
- Méthode de multiplexage faisant preuve de robustesse contre les erreurs
Classification
- CPC, 5
- H04L1/0057
- H03M13/35
- H04J3/0602
- H04L7/10
- H04N19/89
- IPC, 5
- H03M13 35
- H04J3 06
- H04L1 00
- H04L7 10
- H04N19 89
Designated states4
- Contracting states, 4
- Germany
- France
- United Kingdom
- Italy