TWI511499B - Method and system of cross-packet channel coding - Google Patents

Description

Channel coding method and system across packet mode

The present invention generally relates to a channel coding method and system across a packet mode.

The main factors affecting the quality of network transmission can be summarized as: transmission bandwidth, time delay and data loss. The transmission bandwidth must be sufficient to carry the data stream of the application, and the sufficient bandwidth should be provided. When the condition of sufficient bandwidth is unstable, the data stream of the application cannot be completely received and browsed by the user. Therefore, it affects the other two elements of network quality. Time delay In addition to the lack of bandwidth, the various processing times required for data transmission are also influencing factors. Too long time delays can affect the user's willingness to access data and cause poor application quality, such as multimedia applications. The quality of the watch is paused or interrupted. Loss of data, due to the diversification of the network, the loss of packets due to insufficient bandwidth and data errors caused by poor communication environment, the result is that users cannot obtain complete information to obtain information.

Since today's networks do not have a quality assurance mechanism, channel coding techniques are used to provide reliable data transmission in a low-latency manner. Typical examples of channel coding applications include traditional storage devices such as optical discs. Wireless communication protocols such as 802.11 Wireless LAN and 802.16 WiMAX, and video streaming services such as video conferencing. The core technology of channel coding is Forward Error Correction (FEC), which encodes the original data out of the FEC data at the transmitting end and then transmits the two together. The receiving end can tolerate data errors according to the status of data receiving. The decoding action is performed within the number to completely reply the original data. There are two types of channel coding modes known at present: single packet mode and cross packet mode.

Single packet mode to merge original data with FEC encoded data In a single transmission packet, the time delay is low, but generally only the bit error in the data packet can be processed, and the lost data of the packet cannot be recovered. The traditional cross-packet mode channel coding technology is shown in FIG. 1. The FEC codec is that the whole packet is a codec unit, and multiple original packets jointly encode multiple FEC packets, and the source packet and the FEC coded data respectively constitute a transmission packet. Therefore, the data loss problem of packet loss and bit error can be handled at the same time, but the time delay is high due to the large amount of data of the codec and the waiting time of the receiver to collect all the transmission packets. In addition, it is extremely uneconomical to use a packet with a large amount of data to recover a small amount of data, which is extremely uneconomical in the use of bandwidth, so that under a limited transmission bandwidth, the available bandwidth of the original data is often compressed. Not enough to provide users with a high-quality application experience, or the same problem encountered in a single packet mode, channel encoding performance is limited by the length of the network's maximum transmission unit (MTU), which is the FEC packet. The number will be limited and the FEC protection will be reduced.

In general, cross-packet mode is widely used in various networks. Use the service. But with the diversity of the network, the channel coding must be able to adapt to various Changes in the network environment and the accumulation of transmission conditions, such as the amount of data loss may accumulate the final result from different transmission factors, so the low bandwidth usage problem is one of the goals to be improved in cross-packet mode channel coding.

According to an embodiment of the disclosure, a channel coding method across a packet mode is provided. This coding method embodiment includes at least the following steps. The at least one source block packet is split into a plurality of source sub-blocks according to a current sub-block length. The number of required forward error correction code (FEC) sub-blocks is calculated to achieve a desired response rate for a source sub-block. The sub-block is used as a channel coding unit, and FEC encoding is performed on multiple source sub-blocks to generate a plurality of FEC sub-blocks. And the method comprises the sub-block as a constituent unit of the packet assembly, and the plurality of source sub-blocks and the plurality of FEC sub-blocks are assembled into a plurality of transport packets in a cross-packet mode for transmission.

According to an embodiment of the present disclosure, a channel coding system in a cross-packet mode is provided. The coding system embodiment at least includes: a sub-block control unit, disassembling at least one source block packet into a plurality of source sub-blocks according to a current sub-block length, and calculating a plurality of forward error correction codes required The number of (FEC) sub-blocks to achieve the expected recovery rate of a source sub-block. An FEC coding unit performs FEC encoding on the plurality of source sub-blocks by using a sub-block as a channel coding unit to generate the plurality of FEC sub-blocks. And including a packet encapsulating unit, the sub-block is a component unit of the packet assembly, and the plurality of source sub-blocks and the plurality of FEC sub-blocks are assembled into at least one transport packet for transmission.

In order to better understand the above and other aspects of the present invention, a few embodiments are described below, and in conjunction with the drawings, the detailed description is as follows:

301, 302, 303, 304‧‧‧ process steps

401‧‧‧Sub-block control unit

402‧‧‧FEC encoder

403‧‧‧Package package unit

42‧‧‧ Packet data receiving end

421‧‧‧Unpacking unit

422‧‧‧Subblock combination unit

423‧‧‧FEC decoder

4011, 4012, 4013, 4014, 4015, 4221‧‧ ‧ processing blocks

The drawings illustrate exemplary embodiments of the present invention and, together with the description,

FIG. 1 is a diagram showing an example of a channel coding technique of a conventional cross-packet mode.

FIG. 2 is a diagram showing an example of a channel coding technique in a cross-packet mode according to the present disclosure.

FIG. 3 illustrates an exemplary flow of a channel coding method in a cross-packet mode according to the present disclosure.

FIG. 4 is a schematic diagram of an embodiment of a channel decoding system in a cross-packet mode according to the present disclosure.

FIG. 5 is a schematic diagram of an embodiment of a sub-block control unit, and a schematic diagram of an embodiment of a packet data receiving end.

6 is a schematic diagram of an implementation example of a packet format in a cross-packet mode according to the present disclosure.

Figures 7, 8, and 9 show the experimental results: the difference of the playable picture rate in the case of packet loss rate of 1%, 2%, and 3%, respectively.

FIG. 10 is a schematic diagram showing the total bandwidth usage amount obtained by the three experiments of FIGS. 7, 8, and 9.

To further clarify the above and other features and advantages of the present invention, the following detailed description of the embodiments and the accompanying drawings. The above general description and the following detailed description are intended to be illustrative, and are intended to provide a detailed explanation of the disclosure as claimed.

According to an embodiment of the disclosure, a channel coding method and system for a cross-packet mode are disclosed. Referring to FIG. 2, the implementation of the disclosure is disclosed. The sub-block structure used in the example can protect the most common bit errors and packet loss two data loss situations with less coding data, and the packet structure and coding unit used are different from the traditional cross-packet mode. Therefore, it is possible to take into account both the bandwidth usage efficiency and the data recovery performance.

According to the implementation example of the technology of the present invention, the coding operation can be generally performed. It is divided into two parts: firstly, the original data packet is disassembled into a smaller source sub-block, the sub-block is used as a channel coding unit, and the required network transmission condition according to the target application can be selected to generate the required Encoding the sub-block; the second part is to assemble the sub-block into a transport packet, and the assembling method may combine the source sub-block and the coding sub-block to form a trans-packet mode transport packet, or may source the sub-block and code The two sub-blocks of the sub-block are independently assembled into transmission packets across the packet mode, and the two different assembly methods allow the coding sub-block to span several transmission packets. By combining sub-block coding and packet assembly across packet modes, in addition to maintaining the performance of the original cross-packet mode, it is possible to achieve better bandwidth usage efficiency and possibly not limited to the maximum transmission unit of the network.

FIG. 3 illustrates a channel coding method according to the cross-packet mode of the present disclosure. An example of the implementation of the law. In step 301, at least one source block packet is split into a plurality of source sub-blocks according to a current sub-block length. In step 302, the number of required forward error correction code (FEC) sub-blocks is calculated to achieve a source sub-block desired recovery rate. Step 303: FEC encoding the plurality of source sub-blocks by using the sub-block as a channel coding unit, and generating a plurality of FEC sub-blocks, wherein step 303 is based on the plurality of FEC sub-areas required according to step 302. The number of blocks produces multiple FEC sub-blocks. And in step 304, the sub-block is used as a component of the packet assembly, and the plurality of source sub-blocks and the plurality of FEC sub-blocks are The cross-packet mode is separately assembled into a plurality of transport packets for transmission.

One of the channel decoding systems of the cross-packet mode according to the present disclosure As an example, a schematic diagram as shown in FIG. 4 is provided. The coding system includes at least a sub-block control unit 401, which disassembles at least one source packet into a plurality of source sub-blocks according to a current sub-block length, and calculates a plurality of forward error correction codes (FEC) required. The number of blocks to achieve the expected recovery rate of a source sub-block. The encoding system further includes an FEC encoding unit 402 that performs FEC encoding on the plurality of source sub-blocks by using sub-blocks as channel coding units to generate a plurality of FEC sub-blocks. And including a packet encapsulating unit 403, wherein the sub-block is a component unit of the packet assembly, and the plurality of source sub-blocks and the plurality of FEC sub-blocks are respectively assembled into at least one transport packet for transmission.

According to an embodiment of the disclosure, the packet data receiving end 42 can be The network transmission information of the previous sub-block transmitted by the transmitting end is fed back to the sub-block control unit 401 according to the received packet data, and the sub-block control unit 401 can further update the information according to the network transmission of the previous sub-block. The current sub-block length. According to an embodiment of the present technology, the sub-block control unit 401 can update the current sub-block length according to the network transmission information of a previous sub-block. In another embodiment, the sub-block control unit 401 can calculate a one-bit error rate according to a sub-block error rate included in the network transmission information of the previous sub-block, and calculate and update the current sub-area according to the calculation. Block length. In another embodiment, the sub-block control unit 401 can calculate the number of required FEC sub-blocks according to the network transmission information of a previous sub-block. In still another embodiment, the calculation may be performed according to the calculated bit error rate, the updated current sub-block length, and a packet loss rate included in the network transmission information of the previous sub-block. Multiple FEC sub-blocks are required Quantity to reach the expected response rate of the source sub-block.

One of the packet data receiving ends 42 may implement an example, including at least A decapsulation unit 421 is configured to disassemble the transport packet into sub-blocks. The packet data receiving end 42 includes a sub-block combining unit 422 for filtering out the sub-blocks in which the error occurred, collecting the completely received sub-blocks, and reporting the network transmission information of the received data to the sub-blocks of the transmitting end. Control unit 401. The packet data receiving end 42 may further include an FEC decoder 423 for performing a sub-block decoding process to reply to the original data.

There are many possibilities for determining the length of a sub-block and the number of FEC sub-blocks. Embodiments, according to a system implementation example of the present technology, the parameters and the calculation model can be changed according to the network environment in which the target application is located, according to different network transmission conditions, so that the use can be better. FIG. 5 is a schematic diagram showing the processing of an embodiment of the sub-block control unit 401 and the processing of a sub-block combining unit 422 included in an embodiment of the packet data receiving end 42. The calculation of the package encapsulation, and the following description is only a preferred embodiment of the present technology.

In processing block 4011, assuming that the current sub-block length is L bits, and the sub-block error rate due to a bit error is P _B , the bit error rate P _b can be derived, as in equation (1). Illustrative. In an embodiment, the sub-block error rate P _B can be fed back from the sub-block combining unit 422.

Processing in block 4012, based on this bit error rate P _b, can be calculated preferred sub-block length (L _opt), and accordingly updates the current sub-block length. In an embodiment, assuming that the length of the control header required by the sub-block is d- bit, the transmission performance of the sub-block may be defined as E, the L is differentiated, and the differential result is set to 0. L _opt is obtained as shown in the formula (2).

In the processing block 4013 embodiment, the sub-block control unit 401 can derive the new sub-block error rate P _{B_new} using the sub-block length L _opt and the bit error rate P _b . That is, the pushback predicts the sub-block error rate ( P _{B_new} ) that can be caused by the newly obtained sub-block length L _opt . For example, in an embodiment, it is exemplified according to formula (3).

By adding P _{B_new} to the packet loss rate P _{C of the} sub-block due to packet loss, as shown in equation (4), the total loss rate P _{total of} all sub-blocks can be obtained. The embodiment is shown as processing block 4014. In an embodiment, the packet loss rate P _C may be fed back to the sub-block combining unit 422.

P _total = P _{B_new} + P _C (4)

Assuming that the current number of source sub-blocks is k and the expected source sub-block recovery rate is R, then in the embodiment of processing block 4015, the number of FEC sub-blocks h can be calculated by equation (5), and the total sub-area obtained is obtained. The number of blocks is thus (k+h). The condition that the source data can be replied is that the total number of sub-blocks received by the FEC decoder is greater than or equal to the number of source sub-blocks k; for example, k=8, P _total = 0.2, R=0.95, and the input equation (5) can be obtained. The minimum value of h is 5.

Referring to FIG. 4 and FIG. 5, the sub-block combining unit 422 receives the sub-block sent by the de-encapsulated unit, checks the receiving status of the sub-block, and transmits network information (P _B , P _C ) to the packet transmitting end. (Processing block 4221). The sub-block combining unit 422 determines that if the number of sub-blocks is sufficient to reply to the original data, it is transmitted to the FEC decoder. Otherwise, the FEC decoding is skipped and the remaining complete sub-blocks are transmitted.

There are two cases to discuss the number of sub-blocks required for the packet. The first type of FEC sub-blocks constitutes a separate transport packet. Assume that the size of the transport packet is S bits, and the number of sub-blocks per packet is N. , so the number of FEC transport packets is The size of the last FEC transport packet may be less than S. The second type is that the FEC sub-block and the source sub-block together form a transport packet. It is assumed that the maximum transmission unit of the network is MTU, and the number of sub-blocks per MTU is N. , so the number of transport packets is The size of the last transport packet may be less than the MTU.

An implementation example of the encapsulation format of the source packet and the FEC packet can be divided into a header field and a data field as shown in FIG. 6. The header field may include a data header 501, a sub-block header 502, and an FEC header 503. The sub-block header may include an Amount field: record the number of sub-blocks stored in the packet, n, and multiple checksum fields: record the first to n sub-blocks respectively. Checksum value, this checksum value can be used to check if the subblock has an error. Multiple source sub-block data or FEC sub-block data are placed in the data field. The data header and the FEC header are compatible with the traditional cross-packet header, so no detailed description is given here.

On the other hand, combining the FEC sub-block with the source sub-block In an embodiment of a transport packet, the encapsulation format of the transport packet is substantially the same as the example shown in FIG. 6, wherein the number of sub-blocks stored in the packet can be recorded in the Amount field of the sub-block header. The sub-block header may further include an FEC sub-block number field for recording the number of FEC sub-blocks. The source sub-block and the FEC sub-block may be placed in the data field, wherein the source sub-block may be placed before the FEC sub-block. In an embodiment of the present invention, by subtracting the total number of sub-blocks from the number of FEC sub-blocks, it can be known which block of the data field starts as the content of the FEC sub-block.

According to the efficacy of the method and system of the present technology, it is better to achieve better results than previous methods in a plurality of experiments, and the display has an improvement effect, and the experimental results can prove that the case has obvious progress. The experimental environment for verification is as follows: set the network packet loss rate to 1%, 2%, 3%; the average packet loss number is set to 3; the bit error rate is set to 10 ^-1 , 10 ^-2 , 10 ^-3 , 10 ^-4 , 10 ^-5 , 10 ^-6 ; according to 1%, 2%, 3% packet loss rate, the available bandwidth of the network is 224KB / s, 145KB / s, 110KB / s; fixed packet length is 1000 bytes (Bytes); and the original source is compressed into MPEG4 QCIF format "Foreman" movie, the number of frames per second is 30.

The experimental comparison object is mainly compared with the traditional cross-package mode. The performance parameter uses the playable frame rate of the movie, which is defined as assuming that the current number of playable frames per second is f, then the playable picture rate is (f/30). Therefore, the higher the playable picture rate, the more the source data representing the image, the better the response of the FEC. Defined by the National Television Standards Committee (NTSC) The number of frames per second is 30. Another common standard, the Phase Alternation Line (PAL), defines 25 frames per second, which translates to a playable frame rate of 1 and 0.8 respectively.

There are four results in the experiment: the first three pens are the observed packet loss rate. The difference in playback rate can be played in different situations of 1%, 2%, and 3%, as shown in Figures 7-9. The fourth is to calculate the total bandwidth usage of the first three experiments, and is presented in Figure 10.

Experiment 1 is an environment setting with a packet loss rate of 1%. The result is shown in Fig. 7. The method of the invention can obtain a higher playable picture rate under a transmission bit error condition greater than 10 ^-3 ; as for the bit error A rate of 10 ^-1 usually represents a very poor transmission environment, and almost all transmission packets are subject to errors, so performance variability is not observed.

Experiment 2 is an environment setting with a packet loss rate of 2%. The result is shown in Fig. 8. The method of the invention can obtain a higher playable picture rate under a transmission bit error condition greater than 10 ^-3 ; Compared, the loss rate of the packet in the transmission environment increases and the available bandwidth decreases. Under the condition that the bit error rate is 10 ^-2 , the performance improvement provided by the method can be observed to be larger, and the playable picture rate of the method can reach 0.8. At this time, the number of frames per second is 25, and the traditional cross-packet mode compared has only a playable picture rate of about 0.1, and the number of frames per second is only about 3.

Experiment 3 is an environment setting with a packet loss rate of 3%, and the result is as shown in FIG. 9. The method of the present invention can start to obtain a higher playable picture rate and improve under a transmission bit error condition greater than 10 ^-4 . The magnitude increases as the bit error rate increases.

The performance parameter of Experiment 4 is the data usage (overhead), which is determined Meaning Overhead=packet header data size of all transport packets + data volume of all FEC subblocks. The lower the overhead, the higher the efficiency of the bandwidth used by the FEC. Figure 10 shows the difference in data usage between the two comparison methods. The source of the data comes from statistical experiment 1 to experiment 3. The data loss under the conditions of packet loss rate of 1%, 2% and 3% respectively. .

According to the method and system of the present disclosure, the data usage is more than that of the traditional cross The packet mode is less; in fact, in the previous three experimental results, it can be seen that there is better transmission performance than the traditional cross-package mode. According to the disclosure, the source data can be completed with the saved FEC number. Reply, so experiments have shown that better performance can be achieved with limited bandwidth.

In summary, although the present invention has been disclosed above by way of example, it is not intended to limit the present invention. Those who have ordinary knowledge in the technical field of the present invention can make various changes and refinements without departing from the spirit and scope of the present case. Therefore, the scope of protection of this case is subject to the definition of the scope of the patent application attached.

301, 302, 303, 304‧‧‧ process steps

Claims (16)

A channel coding method for a cross-packet mode includes: disassembling at least one source block packet into a plurality of source sub-blocks according to a current sub-block length; and calculating a plurality of forward error correction code (FEC) sub-regions required The number of blocks to achieve a desired recovery rate of a source sub-block; FEC encoding the plurality of source sub-blocks by using the sub-block as a channel coding unit, generating the plurality of FEC sub-blocks; and sub-blocks For the constituent unit assembled by the packet, the plurality of source sub-blocks and the plurality of FEC sub-blocks are assembled into a plurality of transport packets in a cross-packet mode for transmission. The encoding method of claim 1, wherein the calculating comprises: calculating the number of the plurality of FEC sub-blocks according to network transmission information of a previous sub-block. The encoding method of claim 1, wherein the recombining comprises: updating the current sub-block length according to network transmission information of a previous sub-block. The encoding method of claim 1, wherein the calculating comprises: calculating a one-dimensional error rate according to a sub-block error rate included in the network transmission information of a previous sub-block, and updating the The current sub-block length. The encoding method of claim 4, wherein the multiplexing comprises: according to the bit error rate, the updated current sub-block length, and a packet included in the network transmission information of the previous sub-block; The loss rate is calculated by calculating the number of the plurality of FEC sub-blocks to achieve the expected recovery rate of the source sub-block. The encoding method according to any one of claims 2 to 4, wherein the network transmission information of the previous sub-block is fed back from a packet data receiving end. The coding method according to claim 1, wherein the plurality of source sub-blocks and the plurality of FEC sub-blocks are assembled into a plurality of transmissions in a cross-packet mode by using a sub-block as a constituent unit of the packet assembly. The step of the packet further comprises: separately assembling the plurality of source sub-blocks and the plurality of FEC sub-blocks into different transport packets to become the plurality of transport packets. The coding method according to claim 1, wherein the plurality of source sub-blocks and the plurality of FEC sub-blocks are assembled into a plurality of transmissions in a cross-packet mode by using a sub-block as a constituent unit of the packet assembly. The step of the packet further comprises: assembling the plurality of source sub-blocks and the plurality of FEC sub-blocks into the same transport packet to become the plurality of transport packets. A channel coding system for a cross-packet mode, comprising: a sub-block control unit, disassembling at least one source block packet into a plurality of source sub-blocks according to a current sub-block length, and calculating a plurality of forwards required The number of error correction code (FEC) sub-blocks is used to achieve a desired recovery rate of a source sub-block; an FEC coding unit, which uses a sub-block as a channel coding unit, performs FEC encoding on the plurality of source sub-blocks, and generates And the plurality of FEC sub-blocks; and a packet encapsulating unit, wherein the sub-blocks are constituent units of the packet assembly, and the plurality of source sub-blocks and the plurality of FEC sub-blocks are assembled into at least one transport packet for performing transmission. The coding system of claim 9, wherein the sub-block control unit transmits information according to a network of a previous sub-block. Calculate the number of the multiple FEC sub-blocks. The encoding system of claim 9, wherein the sub-block control unit updates the current sub-block length according to network transmission information of a previous sub-block. The encoding system of claim 9, wherein the sub-block control unit calculates the one-bit error rate according to a sub-block error rate included in the network transmission information of a previous sub-block. And according to the update of the current sub-block length. The encoding system of claim 12, wherein the sub-block control unit according to the bit error rate, the updated current sub-block length, and the network transmission information of the previous sub-block A packet loss rate included in the calculation calculates the number of the plurality of FEC sub-blocks to achieve the expected response rate of the source sub-block. The encoding system of any one of claims 10 to 12, wherein the network transmission information of the previous sub-block is fed back from a packet data receiving end. The encoding system of claim 9, wherein the packet encapsulating unit separately assembles the plurality of source sub-blocks and the plurality of FEC sub-blocks into different transport packets to form a plurality of transmissions. Packet. The coding system of claim 9, wherein the package includes: the plurality of source sub-blocks and the plurality of FEC sub-blocks Merging and assembling the at least one transport packet.