JP4043952B2 - Method and apparatus for transferring data packet in communication network - Google Patents

Description

  The present invention relates to a method and apparatus for transferring data units from a transmitter to a receiver. In particular, the present invention relates to compensating for missing or out-of-order data units received by a transcoder.

  Recently, systems and solutions have been developed to provide packet-based transmission of digitally encoded information for various services such as telephones, videophones, television video distributors, as well as Internet surfing and email. Yes. These services are susceptible to delays in various ranges and are therefore often classified to select an appropriate transmission mechanism. Applications that are most sensitive to delay, such as telephone and videophone, may be referred to as real-time services that require a total transmission time of about 200 milliseconds or less from the transmitter to the receiver. Applications that are not sensitive to delay, such as downloading home page data from the Internet, are called best effort services and a transmission time of a few seconds is usually acceptable. When transmission means are shared in this manner, real-time data is generally given priority over best effort data.

  The information to be transmitted is digitally encoded on the transmitter side and matched to a data unit of a specific format, such as a data packet or a frame, according to a code system. A data unit is treated as a transmission unit along a transmission path including, for example, various networks, switches, gateways, and routers. Since the encoded information is displayed or reproduced to the user at the receiving end, it is finally decoded when received at the transmission destination. However, due to inconsistencies in the coding standards, the transmitting and receiving devices can often only use specific coding schemes that are different from each other, before the information encoded by the transmitter reaches the receiving party. Needs to be recoded at a certain location. This operation is often called transcoding.

  Moreover, different networks and / or switching nodes along the transmission path may also use different standard formats and coding schemes, resulting in the need for further transcoding. In this way, the transmission data may have to be recoded several times before reaching the final receiver. It is desirable to minimize delays due to transcoding activity within the nodes of the entire transmission path, both for real-time services and for some media streaming services. For example, if a delay exceeding 200 milliseconds is introduced in a voice call between two parties, the flow of interaction is significantly hindered.

  FIG. 1 illustrates an exemplary communication overview for transferring data between a transmitter 100 and a receiver 102. In this simplified embodiment, asymmetric transmission in only one direction at a point in time is considered. For symmetric transmission, it is of course possible that each party 100 and 102 can be both a transmitter and a receiver, such as a voice call. The transmitter 100 is connected to an access network 104, which may be a local area network (LAN), a wireless network, or the like. Similarly, the receiver 102 is connected to another access network 106. Each access network 104, 106 includes one or more public networks, such as the Internet or a public switched telephone network (PSTN), and is further connected to a respective gateway switch 108, 110 for communicating with the intermediate communication system 112. Is done.

  The transmitter 100 can use the first code system to encode data to be transmitted. On the other hand, the receiver 102 can use the second code system to decode the received data, and the second code system is different from the first. Further, the intermediate communication system 112 may use one or more additional code schemes where the code schemes are different to transfer the communicated data. In this way, it is clear that each time the communicated data enters a new domain that uses a different coding scheme than that used so far, it must be recoded.

  For example, transcoding must be performed between a packet switched network and a circuit switched network using, for example, time division multiplexing (TDM). In this way, data is transferred by one or more transcoders to create a new data unit from the previous data unit before being further transmitted in the transmission path. Gateways 108, 110 as well as other switching nodes in the intermediate communication system 112 of FIG. 1 may generally include such transcoders.

  Further, if one or more networks 104, 112, 106 use packet-based transmission, there is a potential risk that data packets will be received at the transcoder in the wrong order. With packet-based transmission, each data packet may be processed individually and may take a separate transmission path and / or experience different delays at intermediate nodes. Thus, the buffer is often located in the transcoder and receiver to compensate for packets received with different delays and is sometimes referred to as a jitter buffer. The jitter buffer allows reception of a plurality of data packets that are arranged in order before being transcoded or decoded. The packets are then reordered if necessary. Another method is simply to discard packets received in the wrong order, but this results in quality degradation. The receiver side of the intermediate node between the packet domain and the TDM domain generally includes a jitter buffer.

  Jitter buffer technology has the disadvantage of incurring unnecessary delay in transmission. It also requires a clock timing reference when TDM based transmission is involved. Further, the final receiver 102 generally includes a jitter buffer or similar decoding function to increase the total latency of the transmitted data. The transcoding function should not add more latency than is necessary to process the data and perform the transcoding. However, if every data packet arrives sequentially, the resulting quality is degraded when the data packets arrive in the wrong order. In particular, many coding schemes rely on historical data when decoding special large amounts of data. That is, information from previously decoded data is used to decode the next data. Therefore, all data must be decoded in the same sequence order as that output from the transmitting encoder in order to obtain high quality at the receiving end.

  It is an object of the present invention to overcome the disadvantages outlined above and to achieve high quality of the received encoded data while at the same time minimizing the transmission delay in the intermediate transcoder when decoded at the receiving end. It is.

  These objectives are achieved by a simple solution to minimize delay in the transcoder and eliminate the need for jitter buffers, which allows data units that arrive in the wrong sequence order to be Compensated by recovering late arriving data.

  A method and transcoder unit are provided for transcoding communication data in a transmission path between a transmitter and a final receiver. A data unit containing coding information in the first format is received and coded into a new data unit in the second format based on the current coding state extracted from the previous coded data. A new current coding state is extracted each time a data unit is coded.

  As each data unit is received, it is detected whether the received data is in the wrong sequence order, which can be done by reading information indicating the sequence order of the received data. If this is the case, the coding state extracted from the previous coded data is saved and a new pseudo-coded state is extracted to code the next received data unit.

  If the missing data is finally received, it is encoded in the second format based on the stored coding state, and the coded missing data can be used as valid data. The saved coding state can then be deleted. If the corresponding missing data is not received within a preset time period, the stored coding state can also be deleted, but it can be based on data flow characteristics. If the number of stored decode states reaches a preset threshold, the stored coded states can also be deleted, but it can be based on data flow characteristics as well.

  The inventive procedure can be used to decode a received data unit, where the coding state is the decoding state and the second format is the intermediate format. The received data unit may then be decoded into a stream of intermediate data that is divided into intermediate data segments that are encoded into new data units.

  The inventive procedure can also be used to encode a received data unit, where the encoding state is the encoding state and the first format is the intermediate format. The received data unit may then include decoded intermediate data that is encoded into the new data unit.

  If the received data is in the wrong sequence order, error correction data can be derived from the previously received data unit and the current decoding state, where the new pseudo-coding state is based on the derived error correction data. Can be extracted.

  The inventive procedure can be implemented in a transcoder unit located in the transmission path between the transmitter and the final receiver and having means for performing the procedure.

  The inventive procedure may also be implemented in a computer program product that can be loaded directly into a computer in a transcoder unit, including software code means for performing the procedure.

  The inventive procedure may alternatively be implemented in a computer program product stored on a computer usable medium, including a readable program for causing a computer in the transcoder unit to execute the procedure.

  The invention will now be described in more detail with reference to the drawings.

  FIG. 1 illustrates the above exemplary communication system for transferring data between a transmitter 100 and a receiver 102 in which the present invention may be implemented. Of course, the present invention can also be used in many other types of communication systems.

  FIG. 2 is a schematic diagram of the data flow within an exemplary transcoder unit 200 for transcoding data according to the present invention. Data 202 encoded according to a first code scheme and having a first format F is received by a transcoder unit 200 which transcodes the data into a second code scheme having a second format F '. In this embodiment, the transcoding method is divided into two operations, a decoding operation 204 and an encoding operation 206. Between the two operations, the data is in the decoded intermediate format IF 208. However, it is possible within the scope of the present invention to re-code or transcode the first format F input data 202 directly to the second format F 'without using any intermediate format. After the data 210 is encoded in the second format F ′, it is further transmitted to the next node (not shown) in the transmission path. In addition, the transcoder includes a memory 212 that stores the coding state, described in more detail below.

  In general, the input data is grouped into data units, eg packets or frames, encoded according to the first format F. Input data can be packet switched or circuit switched. Each input data unit includes a field having information indicating a sequence order such as a sequence number or a time stamp read by the receiving transcoder unit 200. Normally, if received data units are received in the correct order as originally transmitted and without data loss, the data units are sequentially decoded in operation 204 to the next format IF or F '.

  This embodiment is illustrated in FIG. 3 with further reference to FIG. 2, where F1, F2, F3. . . Are first format F data packets 202 received in the correct order by the transcoder unit 200. Data packet 202 is decoded separately in operation 204 into intermediate data unit 300 having an intermediate format IF. When decoding each data packet 202, information from the previous decoding result shown in FIG. 3 is used as a series of current history decoding states DS302. In addition to the actual data, the decode state includes variables or parameters extracted from the executed decode operation and reflecting some characteristics of the data.

  In this way, the received data packet F1 is decoded 204 into the intermediate data unit IF1 based on the current decoding state DS0 extracted from the decoding result of the previously decoded data packet (not shown). After using decode state DS0, a new decode state DS1 is extracted from the latest decode of F1. The next data packet F2 is then decoded 204 into IF2 based on the last decoding state DS1 and the like.

  In the above description, in practice, the packet length may vary, but it is generally assumed that all input data packets have equal sizes. In that case, the data may be rearranged into equally sized units, but the transcoding method may be handled using the same principles as simplifying equal packet lengths.

  The decoding method illustrated in FIG. 3 relates to data packets received in the correct sequence order. However, if a data packet is received with a higher sequence number that is increased by one over the previous sequence number, it can be assumed that the missing packet has been dropped or delayed. Alternatively, the packet order may be indicated by a time stamp or any other order that indicates information. If data packets are missing in this way, an error correction mechanism can be invoked to minimize the resulting quality degradation at the receiving end. The error correction mechanism may derive error correction data based on the previous data packet and the latest decoding state extracted from the decoding of the previous data packet. Error correction is a well-known procedure that attempts to create the most probable data when real data is not available. In general, error correction is performed according to a predetermined standard of the data format, or a unique mechanism can be used if no standard exists.

  In accordance with the present invention, when data is received in the wrong sequence order, the error correction mechanism is invoked and if the missing data packets are eventually out in the wrong order, error correction. The previous latest decode state is saved for later use. In that case, the missing data packet is decoded according to the present invention based on the stored decoding state.

  FIG. 4 illustrates an exemplary decoding procedure according to the present invention when data packets are received in the wrong sequence order. First, the data packet F1 is decoded 204 into an intermediate data unit IF1 based on the latest decode state DS0, where a new decode state DS1 is extracted as in the previous embodiment of FIG. However, next comes a data packet F3 with a higher sequence number 3 which is increased by one over the previous sequence number 1. Upon detecting that the sequence number of the received data packet F3 is in error, that is, higher than expected, the error correction mechanism 400 derives error correction data 402 in the intermediate format based on the previous decoding state DS1, and creates a new Called to extract the pseudo decode state DS2 ′. Since error correction is a decoder function whose decoding state can be extracted regardless of whether actual data is available, the pseudo decoding state DS2 'can be extracted in the same way as other decoding states.

  The data packet F3 is then decoded 204 into the intermediate data unit IF3 based on the new pseudo decode state DS2 '. Further, the previous decoding state DS1 is stored in the memory 212 in the transcoder unit 200. The new decode state DS3 is also not shown, but is extracted from the decode 204 of the data packet F3 for later use when decoding the next sequenced data packet.

  In this example, the missing data packet F2 may then be decoded 204 into the intermediate data unit IF2 based on the previous decode state DS1 received and stored. The stored decode state DS1 used can then be deleted from the memory 212 in the transcoder unit 200. Finally, the restored intermediate data unit IF2 replaces the previously derived error correction data 402 and is inserted between IF1 and IF3 in the correct sequence order. However, if the missing data packet F2 is not received within a preset time period, for example, when the intermediate format 208 data is encoded 206 in the second format F ′ in the transcoder unit 200, the error correction data 402 is Can remain valid.

  In this way, even if one or more data units in the first format F are received by the transcoder in the wrong order, the effectively decoded intermediate data unit 300 is again in the second format F ′. It can be placed in the correct sequence order before being encoded. Furthermore, this procedure is performed with minimal delay while maintaining the highest possible quality.

  The embodiment of FIG. 4 described above related to one data packet received out of sequence. If multiple data packets are received out of sequence using multiple stored decode states, a corresponding procedure can be performed. In addition, the saved decode state can be deleted from the memory 212 if it remains unused for a preset time period. The preset deletion time limit can be further set or is adaptive depending on data flow characteristics such as the number of data packets in the wrong sequence order that changes over time. Alternatively, the number of decode states stored at the same time can be similarly set depending on the data flow characteristics, or can be maximized by a preset threshold having flexibility. If the number of stored decode states reaches a preset threshold, a mechanism can be activated to delete individual stored decode states from memory 212.

  After the decoding operation 204, the intermediate data unit 300 is arranged in a continuous stream of data 208 that is input to the encoding operation 206. See FIG. Next, depending on the availability of the data, the transcoder unit 200 divides the intermediate data into a new data unit such as a packet, frame or block, ie a data part of a predetermined size, according to the new second coding scheme or format F ′. 208 is encoded. If the intermediate data 208 is not currently available, the encoding method can simply be postponed until it becomes available. The new data unit in the second format F ′ is finally transmitted to the next node in the transmission path.

  FIG. 5 shows the intermediate data units IF1, IF2, IF3. . . FIG. 2 illustrates an encoding procedure according to the invention for an intermediate data stream 208 comprising: First, if intermediate data is received and encoded in the second format F ', new intermediate data segments S1, S2, S3. . . 500. The size of the new intermediate data segment 500 indicated by the vertical dashed line in the figure is different from the size of the intermediate data unit 300 derived from the decoding operation 204 in this embodiment. Each new intermediate data segment or unit 500 is then encoded 206 into a new data unit 506 in the second format F 'based on the current encoding state ES 502 extracted from the previous encoding operation. In addition to the actual data, the encoding state includes variables or parameters that are extracted from the encoding operation being performed and used to encode the data.

  In this way, the intermediate data segment S1 is encoded 206 into a new data unit F'1 based on the last encoding state ES0 extracted from a previously encoded new data packet, not shown. After using the encoding state ES0, a new encoding state ES1 is extracted from the latest encoding of F'1. The next intermediate data segment or unit S2 is then encoded 206 to F'2 based on the last decoding state DS1 and the like.

  When an input data packet or unit in the correct sequence order is not available, the intermediate data unit 300 generates “good data” directly from the payload of the input data packet or unit and “error correction data” is an error correction mechanism. By using 400, it can be classified according to the nature of their content, as derived from the latest decoding state. In the embodiment of FIG. 5, IF1 and IF2 contain "good data", while IF3 contains "error correction data" in the diagram indicated by hatching. Thus, segments S1 and S2 contain only “good data”, whereas segments S3 and S4 result in “good data” and “good data” as a result of the difference in size between intermediate data unit 300 and intermediate data segment 500. Error correction data ".

  According to one aspect of the present invention, the intermediate data segment 500 may be processed differently depending on the classification of the data. In this way, if segment 500 contains at least some “good data”, the segment is encoded and the new encoding state is extracted accordingly. This is true for all segments S1-S4 shown in FIG.

  However, if the entire segment does not contain “good data”, ie only “error correction data”, the segment is simply discarded and at least partially extracted from “good data” The encoding state of is saved. This is illustrated in FIG. 6a, where segment S1 containing “good data” is encoded into a new data unit F′1 and a new encoding state ES1 is extracted. The next segment S2 contains "error correction data" and does not contain "good data", which is indicated by the hatched lines in the figure but is therefore discarded 600. One new data unit is its As a result, it is considered lost. The latest encoding state ES1 is saved for later use if the missing data eventually arrives.

  Referring to FIG. 6b, the segment S1 containing “good data” is encoded into a new data unit F′1, and a new encoding state ES1 is extracted, just as in the embodiment of FIG. 6a. The next arriving segment S2 includes “error correction data” and does not include “good data”, and is again indicated by the diagonal lines in the figure. According to another embodiment, segment S2 is first encoded into pseudo data unit 602 based on available error correction data, which is then discarded 600. In this case, the new pseudo-encoded state ES2 ', not shown, is extracted from the encoding of S2 and used to encode the next segment. In this embodiment, the missing data is finally encoded into a new valid data unit F'2 based on the previously stored encoding state ES1 and arrives later in segment S4. It should be noted that the sequence order of segments S1-S4 is selected from the division of the intermediate data stream for purposes of illustration and should not be confused with the sequence order of actual data. Thus, in this example, segment S4 contains the data that should have been included in segment S2.

  The restored new data unit F'2 can then be transmitted to the next node in the transmission path. In addition, the final receiver can often restore the correct data unit order before decoding, so it is possible to send F′2 after the wrong sequence position, in this case F′3. .

  The procedure described above in connection with FIG. 6b may also be performed for multiple segments arriving out of sequence using multiple stored encoding states. Moreover, the same mechanism as described above for storing multiple decode states can be generally applied to store multiple encode states.

  The embodiment described above includes two separate operations that first decode the received data from the first format to the intermediate format and then encode the intermediate data to the second format for transmission. However, the present invention can also be used to re-encode input data in the first format directly into the second format without using an intermediate format.

  FIG. 7 illustrates the most common case according to the invention of a recoding or transcoding procedure when data units are received in the wrong sequence order. In general, F1, F2, F3. . . Shows an input data unit 700 having a first format F. Data unit 700 has a second format F 'and is generally F'1, F'2, F'3. . . Is encoded separately in operation 702 into a new data unit 704. The expression “encoding” is used here to indicate either decoding, transcoding or encoding. When coding each input data unit 700, information from the previous coding results shown in FIG. 7 is used as a series of current history coding states CS706.

  In this way, the data unit F1 in the first format F is encoded 702 into a new data unit F′1 in the second format F ′ based on the current coding state CS0, where the new coding state CS1. Is extracted. The next received data unit F3 includes a higher sequence number 3 which is increased by one from the previous sequence number 1. Upon detecting that the sequence number of the received data unit F3 is higher than expected, the pseudo data 708 in the second format F ′ is based on the current coding state CS1 to extract a new pseudo coding state CS2 ′. Created. The data unit F3 is then encoded 702 into a new data unit F'3 based on the pseudo-coded state CS2 '. Further, the previous coded state CS1 is stored in the memory 212. A new coding state CS3 is also extracted from the coding 702 of the data unit F3 to a new data unit F'3.

  The missing data unit F2 may then be encoded 702 to a new data unit F'2 based on the previous encoding state CS1 received and stored. The stored coded state CS1 used can then be deleted from the memory 212. Finally, the restored data unit F′2 is used as valid data, and the previously created pseudo data 708 is discarded. However, if the missing data unit F2 is not received within a preset time period, the pseudo data 708 can be used as valid data or discarded, depending on its implementation.

  By using the present invention as described above, the delay in the transcoder is minimized while maintaining the high quality of the transmission data, and the need for a jitter buffer is eliminated. Specifically, data units that arrive in the wrong sequence order are compensated by recovering late-arriving data according to the present invention. The present invention is primarily intended to be used for voice transcoding, but can also be used for video signals based on, for example, fax transcoding, modem data or packets.

  The inventive method may be implemented in software code contained within a computer program product that can be loaded directly into a computer within the transcoder unit 200 or stored on a computer usable medium.

  Although the invention has been described with reference to specific exemplary embodiments, the description is only intended to illustrate the inventive concept and should be taken to limit the scope of the invention Not. Various alternatives, modifications and equivalents may be used without departing from the spirit of the invention as defined by the appended claims.

Schematic diagram of a typical communication system for transferring data Schematic diagram of data flow in the transcoder A logical diagram illustrating the decoding of data packets arriving in the correct order A logical diagram illustrating the decoding of data when data packets arrive in the wrong order A logical diagram illustrating the encoding of intermediate data A logical diagram (a) illustrating the encoding of data when the data unit is dropped and a logical diagram (b) illustrating the encoding of data when the missing data arrives late A logic diagram that generally illustrates the transcoding of data when data units arrive in the wrong order

Explanation of symbols

100 transmitter 102 final receiver 200 transcoder unit 202 data unit 204 encoding 206 encoding 208 intermediate data 300 data unit 302 encoding state 402 error correction data 500 data unit 502 encoding state 506 data unit 700 data unit 702 encoding 704 Data unit 706 Coding state CS2 'Pseudo coding state DS1 Current decoding state DS2' Pseudo coding state ES2 'Pseudo coding state F First format F' Second format F2 Missing data F'2 Encoded shortage Data F3 Receive data unit IF Intermediate format IF2 Encoded missing data S3 Receive data unit S4 Missing data

Claims (18)

-A) receiving a data unit (202, 500, 700) containing coded information in a first format (F, IF);
-B) based on the current coding state (302, 502, 706) extracted from the coding result of the previously coded data unit , in the second format (IF, F '), the received data unit ( 202, 500, 700) are encoded (204, 206, 702) into new data units (300, 506, 704);
-C) Extracting a new current coding state (302, 502, 706) each time a data unit (202, 500, 700) is coded, transmitter (100) and final reception A method for transcoding communication data in a transmission path between the machines (102),
-D) detecting whether the received data is in the wrong sequence order, and if it is in the wrong sequence order,
E) A new pseudo-coding to store the coding state (302, 502, 706) extracted from the previous coding data and to code the next received data unit (202, 500, 700) Extracting the state (DS2 ′, ES2 ′, CS2 ′). -F) receiving the deficient data (F2, S4);
-G) encoding the missing data (F2, S4) into the second format (IF, F ') based on the stored encoding state (302, 502, 706);
-H) using the encoded missing data (IF2, F'2) as valid data.   Method according to claim 2, characterized in that the stored coding state (302, 502, 706) is deleted after step G). If pre SL missing data is not received in the pre-set time period within the stored encoded state (302,502,706), the method according to claim 2, characterized in that it is deleted. 5. The method of claim 4, wherein the preset time period is based on the number of visit data in the wrong sequence order that changes over time .   The stored code state (302, 502, 706) is deleted if the number of stored decode states reaches a preset threshold. Method. 7. The method of claim 6, wherein the preset threshold is based on the number of visit data in the wrong sequence order that changes over time .   The received data unit (202) is decoded in step B), the coding state is a decoding state (302), and the second format is an intermediate format (IF). 8. The method according to any one of 7.   9. The method of claim 8, wherein the received data unit (202) is decoded in step B) into a stream of intermediate data (208) that is divided into intermediate data segments (500).   The method of claim 9, wherein the intermediate data segment (500) is encoded into a new data unit (506).   The received data unit (500) is encoded in step B), the encoding state is an encoding state (502), and the first format is an intermediate format (IF). 8. The method according to any one of 7.   The method of claim 11, wherein the received data unit (500) comprises decoded intermediate data (208) encoded in step B) into a new data unit (506).   The new pseudo-coded state (DS2 ′, ES2 ′, CS2 ′) extracted in step E) is used to code the next received data unit (F3, S3). The method according to claim 1.   14. A method according to any one of the preceding claims, wherein step A) comprises reading information indicating the sequence order of the received data (202, 500, 700, F2, S4).   If the received data is in the wrong sequence order, error correction data (402) is derived from the previous received data unit and the current decode state (302, DS1) and the new pseudo-coded state (DS2). 15. The method according to claim 1, wherein ', ES2', CS2 ') are extracted in step E) based on the derived error correction data (402). -A) means for receiving a data unit (202, 500, 700) containing coded information in a first format (F, IF);
-B) based on the current coding state (302, 502, 706) extracted from the coding result of the previously coded data unit, in the second format (IF, F '), the received data unit ( 202, 500, 700) to (204, 206, 702) encoding the new data unit (300, 506, 704);
C) means for extracting a new current coding state (302, 502, 706) each time a data unit (202, 500, 700) is coded, and a transmitter (100) and final reception A transcoder unit (200) for transcoding communication data in a transmission path between the machines (102),
-D) means for detecting whether the received data is in the wrong sequence order, and if it is in the wrong sequence order,
E) A new pseudo-coding to store the coding state (302, 502, 706) extracted from the previous coding data and to code the next received data unit (202, 500, 700) Means for extracting the states (DS2 ′, ES2 ′, CS2 ′);
A transcoder unit (200) comprising: Flop Rogura beam for the steps according to any of claims 1 to 15, to be executed by the transcoder unit (200). The computer in the transcoder unit (200) includes a readable program for executing the method of any of claims 1 to 15, flop Rogura beam stored in usable media of a computer.

Images