JP4321284B2 - Streaming data transmission apparatus and information distribution system - Google Patents

Description

The present invention is, streaming data transmission apparatus, and an information distribution system.

  Conventionally, in data transmission, there is a method of executing data transmission without going through the procedure of an acknowledgment (ACK) on the receiving side. UDP (User Datagram Protocol) used on IP (Internet Protocol) is a typical example of such a system. In data transmission without such an acknowledgment, there is a problem in that the reliability of data communication is lacking because neither the transmitting side nor the receiving side can detect the absence of transmitted data.

In view of this, there has been considered a method of processing the data itself to be transmitted to improve the reliability of distribution. For example, in Patent Document 1, the data to be transmitted is subjected to FEC (Forward Error Correction) processing to increase the redundancy of the transmission data, and even if the reception side cannot receive a part of the transmission data, the original data from the remaining reception data A technique for restoring the information is disclosed.
Japanese Unexamined Patent Publication No. 2000-78191

  However, in the above-described technique, since data having a predetermined length is subjected to FEC processing, the length of data that can be transmitted cannot exceed the predetermined length. Therefore, data such as streaming data whose length of data to be transmitted cannot be known in advance cannot be transmitted, and Patent Document 1 does not describe a method for solving it.

  The present invention has been made in view of the above points, and it is an object of the present invention to enable streaming data to be transmitted in a data distribution system to which data subjected to FEC processing is transmitted.

The first feature of the present invention for solving the above-described problem is that the streaming data transmitting apparatus buffers the streaming data for a plurality of transmission units, performs FEC processing on the buffered data, and results thereof. The FEC data is transmitted .
In addition, there are a plurality of transmission means for performing transmission. Each of the plurality of transmission means transmits FEC data until buffering of buffering data that is the source of FEC data to be transmitted next is completed. Subsequently, the plurality of transmission units sequentially transmit FEC data for different portions of the same streaming data.

  Thus, in a data distribution system in which data subjected to FEC processing is transmitted, streaming data is buffered for a plurality of transmission units, and FEC processing is performed for each buffered data. Data can be transmitted after FEC processing.

  The “transmission unit” refers to a unit of data transmission such as a frame and a packet.

  Further, “buffering for a plurality of transmission units” means storing until a plurality of units are collected.

  In consideration of performing data transmission wirelessly, for example, when the communicable range of the wireless base station that finally sends data to the receiving side is discretely arranged, the receiving side continuously transmits data. May not be able to receive. Even in such a case, there is a demand for continuous reproduction of time-sequential data such as streaming data.

  The second feature of the present invention is that the streaming data transmitting apparatus wirelessly transmits the FEC data with a transmission capacity higher than the data rate of the buffering data that is the source of the FEC data. Even in an environment where the transmission side cannot receive transmission data continuously, there is a case where streaming data can be reproduced without interruption.

  Note that “buffering data that is the basis of FEC data” refers to buffering data before the FEC data is subjected to FEC processing.

  The data rate refers to the number of bits of data used for reproducing unit time of buffering data.

  A third feature of the present invention is that the transmission means transmits a plurality of FEC data based on different buffering data at the same time. Here, “FEC data based on buffering data” means “FEC data obtained as a result of the processing means performing FEC processing on the buffering data”.

  In this way, compared to a case where only FEC data based on one buffering data is transmitted at a time, a period during which FEC data based on one buffering data can be transmitted is reduced. Can be long.

  In addition, a fourth feature of the present invention is that the transmission means configures a plurality of radio cells along the traveling path of the vehicle equipped with the streaming data receiving device, and simultaneously transmits the same FEC data from the plurality of radio cells. The period from when the vehicle exits one of the plurality of wireless cells to the next wireless cell continues to transmit the FEC data based on one buffering data from the streaming data. This is shorter than the period obtained by subtracting the period for receiving the FEC data for restoring the buffering data.

  Thus, it is possible to prevent missed reception of the FEC data required for restoring the buffering data while the vehicle is traveling between the radio cells. Note that the “period from when one vehicle out of a plurality of wireless cells to the next wireless cell” is calculated using, for example, the average speed of the vehicle on the specific road or the predicted minimum speed can do.

  Further, the streaming data may be streaming data of moving images or music, or may be streaming data in which still images taken repeatedly are continuous. In the case of streaming data in which repeatedly shot still images are continuous, it is preferable that the storage medium be partitioned if the streaming data is buffered by a predetermined number of still images.

  The fifth feature of the present invention is that, among the cyclic identification codes that are cyclically assigned to the buffering data in the order of buffering, the FEC data is assigned to the buffering data that is the source of the FEC data. The cyclic identification code is added and transmitted.

  By using the cyclic identification code in this way, the identification code is repeatedly used in a cyclic manner, so that it is possible to prevent a decrease in data transmission efficiency due to the exhaustion of the identification code or an increase in the data amount of the identification code.

  In addition, the cyclic identification code is a cyclic identification code that returns to the original code when it advances a plurality of times. This multiple times is more than twice the number of buffering data that is the source of the FEC data that the transmission means transmits simultaneously. May be more.

  Further, the set of cyclic identification codes that are cyclically assigned to the buffering data may be different for each type of streaming data that is the source of the FEC data.

  Further, the sixth feature of the present invention is that the streaming data receiving device receives the FEC data transmitted from the streaming data transmitting device as described above, and the FEC data from the received FEC data. Recovery means for recovering the original buffering data, and reproduction control means for performing processing for reproducing streaming data composed of the buffering data based on the data rate of the restored buffering data That is.

  Since such a streaming data receiving apparatus receives the FEC data from the streaming data transmitting apparatus as described above, restores it, and performs processing for reproduction, data distribution to which data subjected to FEC processing is transmitted In the system, the streaming data can be transmitted after FEC processing.

  Further, the restoration means may restore the buffering data that is the basis of the FEC data from the received FEC data based on the cyclic identification code given to the FEC data.

  Further, the restoration means has a plurality of individual restoration means for restoring one buffering data at a time in order to restore a plurality of different buffering data in parallel, and the cyclic identification given to the received FEC data Based on the code, the FEC data may be restored to any one of the individual restoration means.

  The individual restoration means stores the restored buffering data in a common storage medium, and the reproduction control means uses the buffering data based on the time information described in the buffering data stored in the common storage medium. The process for reproducing the streaming data consisting of may be performed.

  Further, the restoration means has M individual restoration means, and the cyclic identification codes assigned to the buffering data restored by the individual restoration means are respectively the cyclic identification code X or the cyclic identification code X. F-1 is provided with an identification code that is M or more ahead of the cyclic identification code X and M prior to the cyclic identification code X. When the receiving means receives the data, the processing of the individual restoring means for restoring the buffering data to which the cyclic identification code X is assigned may be canceled.

  In this way, it is assumed that the FEC data of the current buffering data cannot be received by receiving the FEC data based on the previously advanced buffering data as described above. By canceling, it is possible to prevent the individual restoration means from continuing to restore the FEC data that can no longer be received.

  Also, in a streaming data receiving apparatus that restores one buffering data at a time, the earliest cyclic identification code among the cyclic identification codes assigned to the FEC data received by the receiving means is assigned. The buffering data may be restored. In this way, the streaming data receiving apparatus can restore the FEC data based on the latest buffering data at that time, that is, the buffering data that is most likely to have the longest transmission period thereafter.

  In addition, the restoration unit resets the restoration of the streaming data having the buffering data when the restoration of the buffering data is not completed even after the lapse of a specified time from the start of the restoration of the buffering data by the individual restoration unit. May be. This specified time is a time obtained by subtracting a margin time from a time required for one cycle of the cyclic identification code used in the streaming data transmission apparatus. The margin time is determined by the next ID after one ID is used in the streaming data transmission apparatus. The average value of the time until use is multiplied by twice the number of individual decoding means.

  In this way, the individual restoration means uses the same buffering data as the source of the FEC data to which the same cyclic code based on the separate buffering data is given as a result of the cyclic identification code making a round. It is possible to prevent the data from being restored as if it were FEC data.

(First embodiment)
FIG. 1 shows a configuration diagram of an information distribution system 100 according to the first embodiment of the present invention. The information distribution system 100 includes a streaming server 1, an encoding server 2, wireless LAN access points 4 to 6, and a streaming data receiving device 11 mounted on the vehicle 10.

  Streaming data such as music and images generated by the streaming server 1 is subjected to FEC processing by the encoding server 2 and then converted into FEC data from the wireless LAN access points 4 to 6 via the wide area network 3 such as the Internet. Sent. When the vehicle 10 enters one of the conical wireless cells 7 to 9 formed by the wireless LAN access points 4 to 6, the streaming data receiving device 11 receives the FEC data from the corresponding wireless LAN access point. The streaming data is restored based on the received data, and the streaming data is reproduced.

  FIG. 2 shows a configuration diagram of the streaming server 1, the encoding server 2, the streaming data receiving device 11, and the like.

  The streaming server 1 receives signals such as video and music that are continuously input from the camera 12 and the microphone 13 and changes with time, converts the signals into streaming data, and transmits the signals via the Ethernet (registered trademark) line 51. Output to the encoding server 2. The streaming data refers to data that is continuously transmitted without particularly determining the size of transmission data. Streaming data is often transmitted in a format such as MPEG2 or Real Audio (registered trademark). In this embodiment, the streaming server 1 may be a streaming encoder that is executed on an existing workstation or personal computer that transmits streaming data to the Internet through an Ethernet (registered trademark) line.

  The encoding server 2 buffers the streaming data received from the streaming server 1 for a predetermined time, sequentially encodes each buffered data, and converts the encoded data into a UDP packet, which is wirelessly converted. The data is transmitted into the wireless cells 7 to 9 through the LAN access points 4 to 6.

  Such an encoding server 2 includes a reception interface (referred to as reception I / F in FIG. 2) 21, a CPU 22, a RAM 23 (corresponding to a storage medium in claims), a ROM 24, and a transmission interface (in FIG. 2). 25 (denoted as transmission I / F).

  The reception interface 21 receives the streaming data output from the streaming server 1 to the encoding server 2, converts it into a format that can be recognized by the CPU 22, and outputs the converted data to the CPU 22. With such a function, the CPU 22 can receive data from the streaming server 1 via the reception interface 21.

  The transmission interface 25 processes the data received from the CPU 22 so as to conform to the communication protocol of the wide area network 3 and transmits the processed data to the wireless LAN access points 4 to 6. With such a function, the CPU 22 can transmit data to the wireless LAN access points 4 to 6 via the transmission interface 25 by outputting the data to the transmission interface 25.

  The CPU 22 reads and executes a program from the ROM 24, operates based on the processing contents described in the program, reads information from the RAM 23 and the ROM 24, and writes information to the RAM 23 in the operation. During the operation, the CPU 22 receives streaming data from the streaming server 1 via the reception interface 21, and sends the UDP packet after processing the streaming data to the wireless LAN access points 4 to 6 via the transmission interface 25. Send.

  The wireless LAN access points 4 to 6 are known transmission devices that wirelessly transmit UDP packets received from the encoding server 2 in accordance with the wireless LAN standard. The access points 4 to 6 are arranged along the road so that the radio cells (transmittable ranges) 7 to 9 formed by the access points 4 to 6 are discrete.

  The streaming data receiving apparatus 11 includes a decryption client 14 and a player 15.

  The decoding client 14 receives the UDP packet transmitted from the encoding server 2 via the wireless LAN access points 4 to 6, restores the original streaming data from the received UDP packet, and restores the restored streaming data to the Ethernet (registration). (Trademark) is output to the player 15 via the line 52.

  Such a decryption client 14 has a wireless LAN interface (denoted as wireless LAN I / F in FIG. 2) 41, a CPU 42, a RAM 43, a ROM 44, and a transmission interface (denoted as transmission I / F in FIG. 2) 45. Yes.

  The wireless LAN interface 41 receives the UDP packet transmitted from the wireless LAN access points 4 to 6 via an antenna (not shown), converts it into a format that can be recognized by the CP 42, and outputs it to the CPU 42. With such a function, the CPU 42 can receive data from the wireless LAN access points 4 to 6 via the wireless LAN interface 41.

  The transmission interface 45 processes the data received from the CPU 42 so as to conform to the Internet protocol, and transmits the processed data to the player 15. With such a function, the CPU 42 can transmit data to the player 15 via the transmission interface 45 by outputting the data to the transmission interface 45.

  The CPU 42 reads out and executes a program from the ROM 44, operates based on the processing contents described in the program, reads information from the RAM 43 and the ROM 44, and writes information into the RAM 43. During the operation, the CPU 42 receives a UDP packet from the wireless LAN access points 4 to 6 via the wireless LAN interface 41, and sends the streaming data after restoring the UDP packet to the player 15 via the transmission interface 45. Send.

  The player 15 receives the streaming data from the transmission interface 45 of the decryption client 14, and reproduces the streaming data, that is, displays the video and sound described in the streaming data to the user. As the player 15, a personal computer that reproduces streaming data received via the Internet line, a movie player executed on a workstation, or the like can be used.

  FIG. 3 is a diagram showing a data flow in the information distribution system 100 having such a configuration. The operation of the CPU 22 of the encoding server 2 will be described with reference to this figure.

  First, the streaming server 1 encodes video and music data continuously received from the camera 12 or microphone 13 into a format such as MPEG (step 110). The encoded data is set as streaming data (step 120), and is output to the encoding server 2.

  When receiving the streaming data via the reception interface 21, the CPU 22 of the encoding server 2 buffers the data in the RAM 23. The buffering means that the received streaming data is additionally stored in the RAM 23 until it reaches a certain data amount. Here, the certain amount of data is for a plurality of packets. Multiple packets are the amount of streaming data such that when streaming data is later transmitted as UDP data, the UDP data becomes plural.

  More specifically, the certain amount of data is a certain amount of time for streaming data. The certain amount of time means a data amount that makes the time when streaming data is reproduced constant. Therefore, when the bit rate of streaming data is variable, the data size of the streaming data for a certain time is not constant. Note that the streaming data includes a time stamp (corresponding to the time information in the claims) that specifies which part of the streaming data is to be reproduced at which timing, and the streaming data is based on the time stamp. The time can be specified. In the present specification, buffered data is referred to as buffering data.

  When the buffering of the data for the predetermined time described above is completed, the CPU 22 performs division / encoding (FEC) / encoded packet generation processing on the batch of the buffered data (step 220).

  Then, the CPU 22 collects the data subjected to the above processing into a plurality of UDP packets (step 230), and further transmits the UDP packet via the transmission interface 25 (step 240).

  It should be noted that the CPU 22 newly buffers received data in parallel during the processing of fragmentation / encoding (FEC) / encoded packet generation, and when the buffering is completed, the buffer Steps 220, 230, and 240 are performed on the ring data.

  FIG. 4 is a diagram for explaining the processing of steps 220 and 230 of the CPU 22 as described above. Each rectangle 310-350 in FIG. 4 represents the data processed by CPU22, respectively. The buffering data 310 for a predetermined time stored in the RAM 23 by the process of step 210 is equally divided into a plurality of divided data 311 to 318 by the process of step 220. The number of pieces of one buffering data 310 is determined in advance regardless of the data size of the buffering data.

  Further, the divided data 311 to 318 are encoded by the processing of step 220. Specifically, a known coding matrix such as a Reed-Solomon coding matrix, a Tornado coding matrix, or an LDPC coding matrix is applied to a data string having each of the divided data 311 to 318 as an element. As a result, encoded packets 320 to 335 are generated. The encoded packets 320 to 335 generated in this way have a larger data amount than the original buffering data 310. That is, the buffering data 310 is made redundant by encoding. The encoded packets 320 to 335 made redundant by encoding as described above can be received from the received encoded packets 320 to 335 as long as an arbitrary amount of packets can be received without receiving all of the packets. The buffering data 310 can be restored. Redundancy that can be restored is also called metapacketization.

  Then, by the processing in step 230, some of the encoded packets 320 to 335 are combined into one UDP packet. In FIG. 4, two UDP packets 340 and 350 are shown as an example, but in reality, a larger number of UDP packets are generated. Many UDP packets to be generated may include, for example, randomly selected ones of the encoded packets 320 to 335. When generating the UDP packet 340 and the UDP packet 350, the CPU 22 adds information of IDs 341 and 351 and serial numbers 342 and 352 to the head portion of the packet, respectively, in addition to the encoded packet. The serial numbers 342 and 352 are numbers assigned when the UDP packets are generated so that the original buffering data 310 does not have the same value between the same UDP packets. The IDs 341 and 342 are codes that can identify the original buffering data 310. Therefore, UDP packets based on the same buffering data have the same ID. Each time the buffering data from which the UDP packet is generated becomes new, the ID given to the generated UDP packet is also updated.

  A specific example of the ID update order is shown in FIG. Each of the circles arranged in a circle in FIG. 5 represents one ID. A specific value corresponding to each ID is a number written in the vicinity of the circle. For example, a circle 61 is an ID whose value is 10. Each time the buffering data becomes new, the ID number is updated clockwise as indicated by the arrow in the figure. That is, when 10 is assigned to the UDP data as the ID, the next time the buffering data is updated, 11 is assigned to the UDP packet generated from the buffering data. In this way, the ID is updated as 10 → 11 → 12 →... → 19 as the buffering data is updated. As described above, the ID to be given is a cyclic identification code having a cyclic property.

When such a cyclic identification code is used for an ID, even if streaming data is received from the streaming server 1 one after another and the buffering data is updated one after another, the ID is used repeatedly in a cyclic manner. Decrease in data transmission efficiency due to an increase in the number of digits can be prevented.

  Moreover, the set of IDs to be assigned may be different for each category of the content of streaming data to be transmitted. FIG. 6 shows an example of correspondence between ID groups and streaming data content categories. According to this figure, for streaming data corresponding to the current news category, ID groups 10 to 19 are used, ID groups 20 to 29 are used for sports, and traffic information 30 to 39 is used. ID groups are used, and ID groups 40 to 49 are used for weather forecasting. As described above, when the content category of the streaming data is different for each ID set, the data receiving side selectively receives the UDP packet of the ID corresponding to the category when the streaming data of the specific category is reproduced. That's all you need to do.

  Note that the CPU of the encoding server 2 determines which category of streaming data to be transmitted as a UDP packet belongs, because the information is embedded in the streaming data, and the embedded information is read out. The information may be specified, or the information may be separately received from the streaming server 1.

  FIG. 7 shows the processing timing of the encoding server 2 that receives streaming data from the streaming server 1 and transmits a UDP packet by the operation of the CPU 22 as described above. In the figure, the downward direction corresponds to the direction of time.

In this figure, the streaming server 1 sequentially transmits data S _{1 to} S ₇ having the same reproduction time, which is a unit forming a part of one streaming data. The encoding server 2 receives it and stores these two units (for example, S ₁ and S ₂ ) in the RAM 23 as a fixed time by buffering of the CPU 22 corresponding to step 210 in FIG. Then, processing such as encoding and UDP packetization corresponding to steps 220 and 230 is started, and transmission of a UDP packet is started by the processing of step 240 at the time 65 to 67 when the processing ends.

  The UDP packet transmission is continuously repeated until the next buffering ends 63, 64, and 68. Therefore, the transmission period of the UDP packet based on one buffering data is as shown in a range 69, and after the encoding and UDP packetization of certain buffering data is completed, the buffering of the next buffering data is performed. This is the period until the ring ends.

  Next, the operation after the streaming data receiving apparatus 11 receives the UDP packet transmitted by the encoding server 2 will be described. As shown in the data flow of FIG. 3, the decoding client 14 receives the UDP packet transmitted by the encoding server 2 (step 410), and extracts the encoded packet from the received UDP packet (step 420). Further, the divided data and buffering data are restored from the encoded packet (step 430), and the buffering data is output to the player 15 as streaming data.

  Further, the player 15 decodes the data output from the decoding client 14 such as MPEG (step 510), and reproduces the moving image data, music data, etc. as a result of the decoding (step 520).

  FIG. 8 shows the configuration of a program executed by the CPU 42 of the decryption client 14 in order to realize the data flow as described above. It should be noted that data can be exchanged between programs shown below via a register in the RAM 43 or the CPU 42.

  The CPU 42 executes the reception program 805, the two individual restoration programs 810 and 820, the buffering program 830, and the stream reproduction program 840 in parallel.

  The reception program 805 receives the UDP packet transmitted from the encoding server 2, passes the received UDP packet to the individual restoration program 810 and the individual restoration program 820 as appropriate, and further operates the individual restoration program 810 and the individual restoration program 820. It is a program for controlling. The specific operation of the reception program 805 will be described later.

  The individual restoration program 810 includes a timer program 811, an encoded packet extraction program 812, and a decoding / data restoration program 813 that are respectively executed in parallel by the CPU 42.

  By executing the timer program 811, the CPU 42 measures the elapsed time from the start of the execution, and when the measured time reaches a predetermined expiration time (corresponding to a specified time in the claims), a time-out event Data indicating the occurrence is passed to the reception program 805.

  FIG. 9 is a diagram for explaining the processing of the encoded packet extraction program 812 and the decoding / data restoration program 813. Each rectangle 310 to 350 in FIG. 9 represents data processed by the CPU 42, and the number assigned to each rectangle is the same as the rectangle to which the same number is assigned in FIG. However, the number of received UDP packets 340 and 350 and encoded packets 320 to 335 may be smaller than the encoded packets generated in the process of FIG.

  By executing the encoded packet extraction program 812, the CPU 42 removes the ID, serial number, and the like from the UDP packet (corresponding to the UDP packets 340 and 350 in FIG. 9) passed from the reception program 805, and the encoded packet ( (Corresponding to the encoded packets 320 to 335 in FIG. 9) is extracted, and the extracted encoded packet is passed to the encoded packet extraction program 812.

  By executing the decoding / data restoration program 813, the CPU 42 receives the encoded packets from the encoded packet extraction program 812, and when the received encoded packets reach a predetermined required amount, the encoded packets Is decrypted. Specifically, the decoding is performed by applying an inverse encoding matrix corresponding to the data sequence to the data sequence including each of the necessary amount of encoded packets. As a result, decoded data (corresponding to the divided data 311 to 318 in FIG. 9) is generated. Then, the decoded data is combined to generate buffering data (corresponding to the buffering data 310 in FIG. 9). Further, the generated buffering data is transferred to the buffering program 830, data indicating the occurrence of the decoding completion event is transferred to the reception program 805, and the execution of the individual restoration program 810 is terminated.

  Note that the CPU 42 starts executing the individual restoration program 810 based on performing decoding reservation processing of the UDP packet having a specific ID in the execution of the reception program 805, and at the same time, the timer program 811 and the encoding The execution of the packet extraction program 812 and the decoding / data restoration program 813 is started. Then, when the reception program 805 passes the data indicating that the reservation is canceled to the individual restoration program 810, the CPU 42 receives the data from the individual restoration program 810, and then the timer program 811, the encoded packet extraction program 812. Then, the execution of the decryption / data restoration program 813 ends together with the individual restoration program 810.

  The contents of the individual restoration program 820, timer program 821, coded packet extraction program 822, and decoding / data restoration program 823 are the individual restoration program 810, timer program 811, timer program 821, and coded packet extraction program, respectively. Equivalent to 822. However, the individual restoration program 810 and the individual restoration program 820 are configured to restore UDP packets having different IDs.

  In FIG. 2, two individual restoration programs are shown. However, two or more individual restoration programs may decode UDP packets having different IDs. Note that the number of individual restoration programs to be executed simultaneously is determined in advance.

  By executing the buffering program 830, the CPU 42 stores a plurality of buffering data passed from a plurality of individual restoration programs such as the individual restoration program 810 and the individual restoration program 820 in a common area of the RAM 43, that is, a buffer. save.

  By executing the stream reproduction program 840, the CPU 42 reads a plurality of buffering data stored in the above-described buffer, and at the timing based on the reproduction time stamp described in the buffering data, The ring data is output to the player 15 via the transmission interface 45.

  Here, the process of the reception program 805 for controlling the individual restoration program 810 and the individual restoration program 820 will be specifically described. FIG. 10 shows a flowchart of the reception program 805. The CPU 42 executes the reception program 805 immediately after the decryption client 14 is activated.

  When the CPU 42 starts executing the reception program 805, first, in step 910, the value of the variable Fr is set to “false”. This variable Fr is a variable that is set to “false” when none of the individual restoration programs described above is executed, and “true” when any of the individual restoration programs is executed.

  In step 915, one UDP packet is received via the wireless LAN interface 41. In FIG. 10, the processes other than step 910 are in the form of a loop, but UDP packets can be received sequentially by repeating this loop.

  Next, in step 920, it is determined whether or not the variable Fr is “true”, that is, whether or not at least one individual restoration program is being executed.

  When Fr is “false”, the process of step 925 is subsequently executed, and a reservation process for the individual restoration program is performed by a predetermined number. At this time, assuming that the predetermined number is M, among the ID sets to be used, M IDs are assigned to each of the M individual restoration programs to be reserved in order from the ID of the received packet. As a result, execution of the M individual restoration programs starts.

  In step 930 following step 925, the value of the variable Fr is set to “true”.

  If Fr is “true” in the determination in step 920, then in step 935, the ID of the received UDP packet is read out, and it is determined whether or not this ID is an ID to be newly received. As to whether the ID of the received UDP packet is a new ID to be received, the ID assigned to the buffering data restored by each individual restoration program is X-1 to M-1 ahead. If the ID of the received UDP packet is M or more ahead from X and is an ID before M before X, it is determined that the ID should be newly received. In other cases, it is determined that the ID is not to be newly received.

  FIG. 11 is a diagram for explaining the determination as to whether or not the ID is to be newly received. The circles arranged in a circle in FIG. 11 indicate IDs as circulation identification codes similar to the circles described in FIG. 5, and the update order of the IDs is indicated by solid arrows in FIG. Street clockwise.

  In FIG. 11, M, that is, the number of individual restoration programs is two. In this figure, M = 2 individual restoration programs are executed at the time of step 935, and IDs (value 12 and value 13) indicated by double circles 71 and 72 in FIG. Shows the case. At this point, it is assumed that the restoration of buffering data with IDs 10 and 11 has already been completed.

  At this time, as long as a UDP packet having an ID of 12 or 13 is received from the encoding server 2, the UDP packet may be restored as it is to the individual restoration program to which the corresponding ID is assigned. However, 14 to 19 IDs in the range indicated by the dotted arrow 73 in FIG. 11, that is, 12 IDs (corresponding to X) are M = 2 or more IDs ahead, and 12 IDs are M + 1 = 3. When an ID that satisfies the condition that it is before the previous ID arrives, the restoration of the buffering data that is currently restored can be stopped, and the buffering data before that can be restored.

  In the case where the number of simultaneous transmissions is M, when a UDP packet of an ID to be newly received as described above is transmitted from the ID of X, the ID of M-1 ahead of X and X is already present. It is considered that the transmission of the UDP packet has already ended. Therefore, in this case, if the restoration of the UDP packet with the ID of X is canceled and the UDP packet with the new ID is restored as described above, the process of the individual restoration program is effectively continued while waiting for the UDP packet that is no longer transmitted. You can prevent it from hanging up.

  If it is determined in step 935 that the ID should be newly received, then in step 940, one of the individual restoration programs currently decoded, more specifically, the oldest ID among the individual restoration programs. The restoration process based on the program that is restoring the buffering data is stopped. Specifically, the data indicating that the reservation is canceled is transferred to the one individual restoration program. Thus, the execution of the individual restoration program that has received the data ends.

  When it is determined in step 935 that the ID is not to be newly received, and following steps 930 and 940, in step 945, it is confirmed whether or not event occurrence data has been passed from the individual restoration program. Specifically, the presence / absence of a timeout event and a decoding completion event received from the individual restoration program after the processing of step 945 in the previous loop is confirmed.

  If a decryption completion event has been received, the next reservation is made in step 950. Specifically, among the individual restoration program that has passed the decryption completion event and the ID assigned to the individual restoration program that is currently being executed, the decoding reservation process for the UDP packet of the ID next to the newest ID is performed. .

  If a time-out event has been received, then in step 955, execution of all the individual restoration programs currently being executed is reset (suspended). Specifically, the reservation cancellation data is transferred to all the individual restoration programs currently being executed. That is, the restoration of the entire streaming data including the buffering data with the ID assigned to the individual restoration program that has passed the timeout event is reset. In step 960, the value of the variable Fr is set to “false”.

  If neither a decoding completion event nor a time-out event is received, and subsequent to steps 950 and 960, the processing of step 915, which is the first of the loop, is executed.

  If a plurality of time-out events and decoding completion events are received at a time, the corresponding steps 950, 955, and 960 may be performed in parallel.

  As processing of the reception program 805, the CPU 42 specifies the ID of the UDP packet received from the encoding server 2 in parallel with the processing described in the flowchart of FIG. 10, and performs individual restoration assigned to the ID. The UDP packet is passed to the program. In this way, the individual restoration program can receive the UDP packet of the ID assigned to itself and use it for the restoration process.

  Here, a predetermined expiration time in the timer program of the individual restoration program will be described. The predetermined expiration time is a time obtained by subtracting the margin time from the time for which the identification code used in the encoding server 2 makes a round.

  The time for which the ID makes a round is the UDP packet based on the buffering data from the time when the encoding server 2 starts to transmit the UDP packet having the ID and the ID of X is assigned to the buffering data. Is transmitted, the ID assigned to the subsequent buffering data is updated, the ID of the X is assigned to the buffering data again, and the time until the UDP packet having that ID starts to be transmitted is Say.

  Further, the margin time is obtained by multiplying the average value of the time from when one ID is used in the encoding server 2 to when the next ID is used by twice the predetermined number of simultaneous executions of the individual restoration program. It was time. The time from when one ID is used until the next ID is used is the time when the next buffering is started from the point when an ID of X is assigned to certain buffering data and a UDP packet having that ID starts to be transmitted. This is the time until the next ID of X is assigned to data and a UDP packet having that ID starts to be transmitted.

  As a result, the decoding client 14 cannot receive any UDP packet from the encoding server 2 for some reason, such as when the vehicle on which the streaming data receiving device 11 is mounted is significantly out of the range of the wireless cells 7 to 9. Even in this case, the execution of the individual restoration program can be reset after the expiration time described above has elapsed. As described above, the expiration time is reached before the ID makes a round, so that when reception is possible again, there is no possibility of receiving UDP packets based on different buffering data with the same ID.

  Through the processing of the reception program 805 as described above, the CPU 42 appropriately restores a plurality of buffering data by passing the sequentially received UDP packets to the individual restoration program 810 and the individual restoration program 820 as appropriate (step 925). If the UDP packet has an ID to be newly received (step 935), the oldest decoding process is stopped and decoding by the UDP packet is started (step 940).

  When the time-out event occurs, the restoration of the streaming data is reset (step 955). When the decoding completion event occurs, the decoding of the buffering data of the next ID is reserved (step 950).

  Further, until the time-out event occurs, the received UDP packet is passed to the individual restoration program to which the ID of the UDP packet is assigned and is continuously decoded.

  By the operation of the data distribution system as described above, the encoding server 2 buffers the streaming data for a plurality of packets, performs encoding on the buffering data, and transmits the resulting UDP packet. . Since it is like this, in the data distribution system in which the data subjected to the encoding process is transmitted, the streaming data is buffered for a plurality of packets, and the encoding process is performed for each buffered data. Streaming data can be encoded and transmitted.

  Note that the UDP packet transmission capacity (transmission data amount per unit time) in the encoding server 2 and the UDP packet transmission capacity in the wireless LAN access points 4 to 6 are the data rate of the buffering data from which the UDP packet is based. It may be a higher value. As a result, the streaming data receiving apparatus 11 can receive the UDP packet of the buffering data in a time shorter than the time taken to reproduce the buffering data itself. Therefore, for example, when the wireless cells 7 to 9 are discretely discarded, there is a case where the streaming data can be reproduced without interruption even in an environment where the receiving side cannot receive the transmission data continuously.

(Second embodiment)
Next, a second embodiment of the present invention will be described. The present invention is different from the first embodiment in that there are a plurality of encoding servers 2.

  The two encoding servers each receive the same streaming data from the same streaming server 1 and perform the same processing as the first embodiment on the received data. However, each encoding server 2 performs buffering, encoding, UDP packetization, transmission, and the like for different portions of the same streaming data. Specifically, each encoding server 2 sequentially performs processing such as buffering, encoding, UDP packetization, and transmission on each portion of the same streaming data equally divided by the data size.

For example, when there are two encoding servers 2, one piece of streaming data is equally divided by the data size, and each equally divided part is processed alternately. Specifically, assuming that each part of the streaming data is S ₁ , S ₂ , S ₃ , S ₄ , S ₅ , S ₆ ..., The first encoding server 2 has S ₁ , S ₃ , then process S _5, a second encoding server 2 performs processing to _{S 2, S 4, S 6} . Such allocation of streaming data is determined in advance, and each encoding server 2 determines which part is to be processed in accordance with the determination, and does not receive or discards an unnecessary part of the streaming data. It is like that. The division of the portion of the streaming data may be equally divided by the data size as described above, or may be equally divided by the reproduction time based on the time stamp of the reproduction time recorded in the streaming data.

  FIG. 12 is a timing chart showing processing timings of the two encoding servers 2 that receive streaming data from the streaming server 1 and transmit UDP packets in the present embodiment. The display format is as shown in FIG. Hereinafter, the first encoding server 2 is referred to as encoding server # 1, and the second encoding server 2 is referred to as encoding server # 2.

In FIG. 12, data S ₁ , S ₂ , S ₃ , S ₄ , S ₅ , S ₆ , S ₇ ..., Which are unit parts of equal size of one streaming data, are transmitted in this order from the streaming server 1. Has been. The encoding server # 1 and the encoding server # 2 are determined in advance so as to alternately process every two unit portions of the streaming data. Specifically, the encoding server # 1 receives the data of the S ₁ and S _2, encoding of collectively as shown in step 210 to 240 in FIG. 3 and buffers, UDP packetized transmission performs processing etc., then the same reference numerals with respect to S ₅ and S _6, UDP packetization, the process of the transmission are performed. The encoding server # 2 receives the data of the _{S 1} and _{S 2,} encoding as shown and buffered collectively to steps 210-240 in FIG. 3, UDP packetization, such as transmission Process.

Note that, as described with reference to FIG. 7 in the first embodiment, the encoding server 2 completes the buffering of the next buffering data after completing the encoding of the buffering data and the UDP packetization. The UDP packet is continuously transmitted many times. In the present embodiment, since the encoding server # 1 and the encoding server # 2 perform processing alternately, the transmission time of the UDP packet for one buffering data is doubled. Specifically, in the encoding server # 1, between the timing 75 to the timing 76, and transmits the UDP packet for buffering data of S ₁ and S _2. In the encoding server # 2, between the timing 77 to the timing 78, and transmits the UDP packet for buffering data of _{S 3} and _{S 4.}

  If all the streaming data is processed by one encoding server # 1, the buffering data of S3 and S4 must be buffered, encoded, and transmitted in the encoding server # 1. The UDP packet transmission time for buffering data is halved.

  As described above, since a plurality of encoding servers transmit different buffering data in order, in one encoding server, after the UDP packet for one buffering data starts to be transmitted, The time until the buffering of the buffering data is completed is extended.

  On the other hand, when transmission data of a plurality of encoding servers are all transmitted to the wireless LAN access points 4 to 6 on the same communication path, the transmission efficiency at the wireless LAN access points 4 to 6 is lowered. However, when the capacity of the communication channel to be used is sufficiently large compared to the data rate of streaming data (that is, the number of bits of data used for reproducing streaming data per unit time), this reduction is not a problem.

  Below, this effect is demonstrated by a simple trial calculation. In the following calculation, for the sake of simplicity, the redundancy that the amount of encoded data increases from the buffering data accompanying the encoding process is not considered.

  For example, when the data rate of streaming data is 160 kbit / s and the data capacity of buffering data is 100 kbytes, data for 5 seconds is held in one buffering data. When this is transmitted through a communication path with a transmission path capacity of 8 Mbit / s, it takes 0.1 seconds to receive one buffering data when the encoding server is used alone. In the case of the present embodiment in which two units are operated in parallel, the time is extended to 0.2 seconds. On the other hand, the time during which the data of each buffer continues to be transmitted is Tm = 5 seconds in the former case, whereas Tm ′ = 10 seconds in the latter case.

  To summarize the above calculation, when the encoding server is used alone, the receiver needs to receive the UDP packet for 0.1 second in 5 seconds. What is necessary is just to receive the UDP packet for 2 seconds.

  As described above, when the transmission time of the UDP packet for one buffering data is extended, the streaming data receiving device 11 of the vehicle 10 traveling between the wireless LAN access points 6 to 8 may possibly lose reproduction of the streaming data. Is reduced. FIGS. 13 and 14 are diagrams for explaining the influence of the extension of the UDP packet transmission time on the possibility of interruption of streaming data reproduction.

  FIG. 13 is a diagram illustrating the relationship between the travel of the vehicle 10 and the transmission time of the UDP packet when the encoding server performs processing alone. FIG. 14 is a diagram showing the relationship between the traveling of the vehicle 10 and the transmission time of the UDP packet when there are a plurality of encoding servers as in this embodiment.

  The vehicle 10 equipped with the streaming data receiving device 11 that passes through the wireless LAN access point 4 and the wireless LAN access point 5 continuously is included in the wireless cell 7 and the wireless cell 8 and the required number of UDPs. A packet needs to be received. On the other hand, the UDP packet for one buffering data is transmitted in Tm seconds in the case of FIG. 13 and in Tm ′ seconds longer than Tm seconds in the case of FIG.

  For example, as shown in FIG. 13, when transmission of a UDP packet of buffering data when the vehicle 10 leaves the wireless cell 7 starts and the vehicle 10 cannot enter the next wireless cell 8 within Tm seconds, the streaming data The receiving device 11 cannot receive the buffering data, and reproduction of that portion of the streaming data is lost. However, if the transmission time of the buffering data is long as shown by Tm ′ in FIG. 14, the possibility that the vehicle 10 has entered the wireless cell 7 or the wireless cell 8 during that time is high. If Tm ′ is longer than the time from the start to the time of entering the radio cell 8, the vehicle 10 enters the radio cell at least once within the time T′m. If the UDP packet of the buffering data continues to be transmitted for 200 or more time required for the vehicle 10 to receive the necessary UDP packet while in the wireless cell, the buffering data is restored and reproduced. It becomes possible, and continuous reception of streaming data becomes possible.

  In this way, if the number of encoding servers is increased so that different buffering data is transmitted at the same time, even if the radio cells are in a discrete arrangement, the vehicle outputs one radio cell among a plurality of radio cells. In order for the decoding client 14 to restore the buffering data from the period from the start to the next wireless cell, that is, the period during which the time between wireless cells continues to transmit a UDP packet based on one buffering data UDP packet necessary for restoring buffering data while the vehicle 10 is traveling between the radio cells if the period is shorter than the period obtained by subtracting the period required to receive the UDP packet. Can be missed.

  In this way, it is also possible to reduce the possibility of lack of reproduction of streaming data by adjusting the interval between discrete wireless cells. In other words, the wireless LAN access point is placed at such a distance that transmission of UDP packets for transmission of buffering data within the time from one wireless cell to the next wireless cell does not end. do it.

  In order to specify the time between the wireless cells described above, it is necessary to specify the distance of the unreceivable area between the wireless cells and the traveling speed of the vehicle. If so, the average speed of the vehicle on the road may be used as the travel speed of the vehicle, and if the reliability is improved and reproduction without loss is guaranteed for most vehicles, the road What is necessary is just to use the minimum traveling speed assumed in (1).

  In order to exhibit the effect of the present embodiment, it is not always necessary to have a plurality of encoding servers 2 as described above. For example, in the CPU 42 of one encoding server 2, steps 210 to 240 in FIG. By performing a plurality of processes in parallel, the same effect as this embodiment is exhibited. That is, it is only necessary to simultaneously transmit UDP packets based on different buffering data.

  Further, in the present embodiment, the ID set assigned to the UDP packet returns to the original when the number of times that is more than twice the number of buffering data simultaneously transmitted by the encoding servers # 1 and # 2 is increased. Such a cyclic identification code is used.

(Third embodiment)
Next, a third embodiment of the present invention will be described. The present embodiment is different from the second embodiment in that the streaming data transmitted from the encoding server 2 is not a moving image or music but a continuously stopped image. Here, the continuous still image refers to streaming data in which one buffering data includes data of one or a plurality of still images. In the present embodiment, one piece of buffering data includes one piece of still image data.

  In the present embodiment, there are three encoding servers instead of two, each of which sequentially processes one still image as buffering data, such as buffering, encoding, UDP packetization, and transmission. I do.

  FIG. 15 shows the UDP packet transmission timings of these three encoding servers. The downward direction in the figure is the direction of time flow, and the solid lines 81, 82, and 83 indicate the transmission timings of the UDP packets of the first, second, and third encoding servers, respectively. Moreover, the black circles on each solid line indicate the transmission start and transmission end timing of the UDP packet for one buffering data.

  For example, at the timing of the black circle 84, the first encoding server starts repeated transmission of the UDP packet based on the buffering data of 1 ID, and at the timing of the black circle 85, the transmission of the UDP packet is terminated. Further, at the timing of the black circle 86, the second encoding server starts to repeatedly transmit the UDP packet based on the buffering data of the ID of 5, and at the timing of the black circle 87, the transmission of the UDP packet is ended. Further, at the timing of the black circle 88, the third encoding server starts to repeatedly transmit the UDP packet based on the buffering data of the ID of 3, and at the timing of the black circle 89, the transmission of the UDP packet is ended. In the present embodiment, 1, 2,..., 9 are used cyclically as ID sets.

  Here, a time Tl from when transmission of a UDP packet for one buffering data is started until transmission of a UDP packet for another buffering data is started in the same encoding server is the reception of data, The processing speed, transmission timing, etc. of the encoding server are set so that the streaming data receiving apparatus 11 having a low speed of decoding, reproduction, etc. has the longest time required for receiving / decoding the buffering data. The longest time is not the longest time including the time required for reception / decoding in the case where the streaming data receiving device 11 is out of order, but in the normal operation of the streaming data receiving device 11. This is the longest time within the fluctuation range of time required for reception and decoding.

  By doing so, even the low-speed streaming data receiving apparatus 11 can reproduce at least one still image. The operation of the low-speed client will be described later.

  In addition, the time Th from when the transmission of the UDP packet for one buffering data to the transmission of the UDP packet for the buffering data of the next ID from another encoding server is started. The processing speed, transmission timing, and the like of the encoding server are set so that the streaming data receiving apparatus 11 that performs high-speed processing such as reception, decoding, and reproduction has the longest time required to receive and decode the buffering data. By doing in this way, the high-speed streaming data receiver 11 can reproduce | regenerate all the still images. As the high-speed streaming data receiving apparatus 11, the streaming data receiving apparatus 11 having the decoding client 14 that executes a plurality of individual restoration programs in parallel as shown in the first embodiment can be considered.

  However, the wireless LAN access points 4 to 6 which are the transmission paths of the UDP packets send the UDP packets transmitted at the same time to the wireless cells 7 to 9 with fine transmission timing allocation by time division. FIG. 16 shows UDP packet transmission timings in the actual wireless LAN access points 4 to 6 by time division.

  In FIG. 16, the downward direction in the figure indicates the direction of time. Each of the rectangles 91 to 102 connected in the center vertically indicates the period of Th in FIG. The numbered rectangles associated with the respective rectangles 91 to 102 are time-division time slots in the period, and the numbers assigned to the rectangles are included in the UDP packet transmitted in the time slot. ID value.

  For example, the period of the rectangle 91 corresponds to the period in which only the UDP packet with ID 1 is transmitted from the first encoding server in FIG. 15, so that in all the time slots at the wireless LAN access points 4 to 6. A UDP packet with an ID of 1 is sent out.

  The period of the rectangle 92 corresponds to the period in which the UDP packet with ID 1 is transmitted from the first encoding server and the UDP packet with ID 2 is transmitted from the second encoding server in FIG. Therefore, a UDP packet with ID 1 is sent in half of the time slot, and a UDP packet with ID 2 is sent in the other half.

  Further, in the period of the rectangle 93, the UDP packet with ID 1 is transmitted from the first encoding server in FIG. 15, the UDP packet with ID 2 is transmitted from the second encoding server, and the third encoding is performed. Since it corresponds to the period in which the UDP packet with ID 3 is transmitted from the server, the UDP packet with ID 1 is sent out in 1/3 of the time slot, and the UDP packet with ID 2 is sent out in the same 1/3 time slot. Similarly, a UDP packet with 1/3 and ID 3 is transmitted. Further, the sending order may be the order of IDs as shown in FIG.

  In this way, when UDP packets for a plurality of IDs are sent out at the same time in the wireless LAN access points 4 to 6, they are sent in a time-sharing manner at the same frequency.

  Next, as an example of the low-speed streaming data receiving apparatus 11 described above, an operation of the decoding client 14 in which the number of simultaneous executions of the individual restoration program executed by the CPU 42 in the decoding client 14 is 1 will be described. The low-speed decryption client 14 differs from the decryption client 14 of the first embodiment as described above in that the number of simultaneous executions of the individual restoration program executed by the CPU 42 is 1, and the reception program 805 is a flowchart of FIG. It is something like that shown in Hereinafter, the operation of the CPU 42 that executes the flowchart of FIG. 15 will be described.

  15 and 10 are the processes for a plurality of individual restoration programs in the case of FIG. 10, except that they are the processes for one individual restoration program in FIG. The same processing is performed, and detailed description thereof is omitted here.

  If it is determined that step 920Fr is “false”, that is, if the decoding process of the individual restoration program is not performed, the process of step 525 is subsequently executed. In step 525, m UDP packets are received via the wireless LAN interface 41. Here, m is the number of samples received to check which ID the latest UDP packet is, and a value equal to or greater than the number of simultaneous transmissions of UDP packets with a plurality of IDs (3 in this embodiment). It is.

  Next, in step 530, a decryption reservation process is performed for the received UDP packet having the newest ID. As a result, the execution of the individual restoration program to which the ID is assigned is started, and when the UDP packet having the ID passed from the reception program 805 reaches the required amount by the execution, processing such as decoding and restoration is performed. I do.

  Next, in step 535, the variable Fr is set to “true”.

  If the variable Fr is “true” in step 920 and after step 535, in step 545, it is confirmed whether or not event occurrence data has been passed from the individual restoration program. Specifically, the presence / absence of a time-out event and a decoding completion event received from the individual restoration program after the processing of step 545 in the previous loop of steps 920 to 960 is confirmed.

  If a decryption completion event has been received, then in step 960, the value of the variable Fr is set to “false”.

  If a time-out event has been received, the decoding process is reset and the variable Fr is set to “false” as in the case of FIG.

  If neither a decoding completion event nor a time-out event is received, and after step 960, the processing of step 910, which is the first of the loop, is executed.

  By executing the processing of the reception program 805 of this embodiment, when the decoding is completed and when the decoding times out, the CPU 42 sets Fr to “false” in step 960, and again. In steps 525 to 535, the latest UDP packet from those that can be received can be decoded and restored.

  As processing of the reception program 805, the CPU 42 receives a UDP packet from the encoding server 2 in parallel with the processing described in the flowchart of FIG. 15, specifies the ID of the received UDP packet, If the ID is currently assigned to the individual restoration program, the UDP packet is passed to the individual restoration program. In this way, the individual restoration program can receive the UDP packet of the ID assigned to itself and use it for the restoration process.

  FIG. 16 illustrates an encoding server in which the low-speed streaming data receiving apparatus 11 that starts decoding the next newest UDP packet when decoding is completed or times out performs transmission as illustrated in FIGS. 15 and 16. An example of the order of UDP packets to be restored from is shown. Each circle in FIG. 16 corresponds to one ID. In the example of this figure, the low-speed streaming data receiving apparatus 11 receives 1, 4 and 7 and skips the buffering data, and decodes and restores it. Therefore, the reproduced still image is skipped by two images.

  In each of the above embodiments, the encoding server 2 (encoding servers # 1 and # 2 in the second embodiment) and the wireless LAN access points 4 to 6 constitute a streaming data transmitting apparatus. However, the wireless LAN access points 4 to 6 are not necessarily components of the streaming data transmission device, and there may be an embodiment in which only the encoding server 2 constitutes the streaming data transmission device.

  Further, the CPU 22 of the encoding server 2 functions as a processing unit by executing the processing of steps 220 and 230 in FIG.

  A UDP packet corresponds to FEC data. However, the FEC data need not be a UDP packet. Any data may be used as long as it is data after the FEC process is executed.

  Further, the transmission interface 25 and the wireless LAN access points 4 to 6 of the encoding server 2 (encoding servers # 1 and # 2 in the second embodiment) correspond to transmission means.

  The wireless LAN interface 41 of the decryption client 14 corresponds to a receiving unit.

  Further, the CPU 42 of the decryption client 14 functions as a restoration unit by executing the reception program 805, the individual restoration programs 810 and 820, and the buffering program 830.

  Further, the CPU 42 of the decryption client 14 functions as a playback control unit by executing the stream playback program 840.

  Further, the CPU 42 of the decryption client 14 functions as individual restoration means by executing the individual restoration programs 810 and 820, respectively.

  In the above-described embodiment, the decryption client 14 and the player 15 are separate, but the decryption client 14 and the player 15 are integrated, and the decryption client 14 and the player 15 have the same processing. It may be realized in the CPU.

  The RAM 43 of the decryption client 14 corresponds to a “common storage medium” in the claims.

It is a lineblock diagram of the information distribution system concerning a 1st embodiment of the present invention. 1 is a configuration diagram of a streaming server 1, an encoding server 2, a streaming data receiving device 11, and the like. It is a figure which shows the flow of the data in such an information delivery system 100. FIG. It is a figure explaining the process of steps 220 and 230 of CPU22. It is a figure which shows an example of the update order of ID provided to a UDP packet. It is a graph which shows a response | compatibility with the group of ID and the category of the content of streaming data. It is a timing chart which shows the timing of the process of the encoding server 2 which receives streaming data from the streaming server 1 and transmits a UDP packet. It is a figure which shows the structure of the program which CPU42 of the decoding client 14 performs. It is a figure explaining the process of the encoding packet extraction program 812 and the decoding and data restoration program 813. 7 is a flowchart showing processing of a reception program 805. It is a figure explaining determination of whether ID of the received UDP packet is ID which should be newly received. It is a timing chart which shows the timing of the process of two encoding server # 1 which receives streaming data from the streaming server 1, and transmits a UDP packet, and encoding server # 2. It is a figure which shows the relationship between the driving | running | working of the vehicle 10, and the transmission time of a UDP packet in case an encoding server processes independently. It is a figure which shows the relationship between driving | running | working of the vehicle 10 in case there exist multiple encoding servers, and the transmission time of a UDP packet. It is a figure which shows the transmission timing of the UDP packet of three encoding servers in 3rd Embodiment. It is a figure which shows the transmission timing of the UDP packet in the actual wireless LAN access points 4-6 by a time division. It is a flowchart which shows the process of the reception program 805 of the low-speed streaming data receiver 11. FIG. It is a figure which shows an example of the order of ID of the buffering data received, decoded, and decompress | restored in the low-speed streaming data receiver.

Explanation of symbols

1 ... streaming server, 2 ... encoding server, 3 ... wide area network,
4-6 ... Wireless LAN access point, 7-9 ... Wireless cell, 10 ... Vehicle,
11 ... Streaming data receiver, 12 ... Camera, 13 ... Microphone,
14 ... Decryption client, 15 ... Player, 21 ... Reception interface,
22 ... CPU, 23 ... RAM, 24 ... ROM, 25 ... transmission interface,
41 ... Wireless LAN interface, 42 ... CPU, 43 ... RAM, 44 ... ROM,
45 ... Transmission interface, 51 ... Ethernet (registered trademark) line, 52 ... Ethernet (registered trademark) line,
100: Information distribution system, 310: Buffering data,
311 to 318 ... divided data, 320 to 335 ... encoded packet,
340, 350 ... UDP packet, 341, 351 ... ID,
342, 352 ... serial number, 805 ... reception program, 810 ... individual restoration program,
811: Timer program, 812: Encoded packet extraction program,
813 ... Decryption / data restoration program, 820 ... Individual restoration program,
821: Timer program.

Claims (8)

A storage medium for buffering streaming data for a plurality of transmission units;
Processing means for performing FEC (Forward Error Correction) processing on buffered data for a plurality of transmission units buffered by the storage medium;
Transmission means for transmitting the FEC data obtained as a result of the FEC processing performed by the processing means so that a streaming data receiving apparatus that restores buffering data from the received FEC data can receive the data.
There are a plurality of transmission means,
Each of the plurality of transmission means continues the transmission of the FEC data until the buffering of the buffering data that is the source of the FEC data to be transmitted next is completed,
The streaming data transmitting apparatus characterized in that the plurality of transmitting means sequentially transmit FEC data for different portions of the same streaming data.   2. The streaming data according to claim 1, wherein the transmission unit transmits the FEC data wirelessly, and a transmission capacity of the transmission is larger than a data rate of streaming data that is a source of the FEC data. Transmitter device.   The streaming data transmitting apparatus according to claim 1, wherein the streaming data is moving image or music streaming data. The streaming data is streaming data in which still images taken repeatedly are continuous,
4. The streaming data transmitting apparatus according to claim 1, wherein the storage medium buffers the streaming data for a predetermined number of the still images. The processing means includes, among the cyclic identification codes that are cyclically assigned to the buffering data in the order of buffering, the cyclic identification code assigned to the buffering data that is the source of the FEC data. Grant,
5. The streaming data transmission apparatus according to claim 1, wherein the transmission unit transmits the FEC data to which the cyclic identification code is added.   The cyclic identification code is a cyclic identification code that returns to the original code when it advances a plurality of times, and this multiple times is twice the number of buffering data that is the source of the FEC data simultaneously transmitted by the transmission means. 6. The streaming data transmitting apparatus according to claim 5, wherein the number of times is greater than the number of times.   7. The streaming data transmitting apparatus according to claim 5, wherein a set of cyclic identification codes that are cyclically assigned to the buffering data is different for each type of streaming data that is the source of the FEC data. An information distribution system comprising a streaming data transmitting device and a streaming data receiving device,
The streaming data transmitting device is
A storage medium for buffering streaming data for a plurality of transmission units;
Processing means for performing FEC (Forward Error Correction) processing on buffered data for a plurality of transmission units buffered by the storage medium;
Transmission means for transmitting the FEC data obtained as a result of the FEC processing performed by the processing means so that a streaming data receiving apparatus that restores buffering data from the received FEC data can receive the data.
There are a plurality of transmission means,
Each of the plurality of transmission means continues the transmission of the FEC data until the buffering of the buffering data that is the source of the FEC data to be transmitted next is completed,
The plurality of transmission means sequentially transmit FEC data for different parts of the same streaming data,
The streaming data receiving device receives FEC data transmitted from the streaming data transmitting device;
Restoring means for restoring the buffering data that is the source of the FEC data from the FEC data received by the receiving means;
An information distribution system comprising: reproduction control means for performing processing for reproducing streaming data composed of the buffering data based on a data rate of the buffering data restored by the restoration means.

Images