JP5532638B2 - Data distribution system - Google Patents

Description

  The present invention relates to a data distribution system for distributing real-time data such as audio data from a source device, which is a data supply source, to a client device via a server device.

  2. Description of the Related Art Conventionally, a system that distributes data such as voice data from a server device to a client device is known. For example, in the recorded data distribution system of Patent Document 1, although not real-time data, the recorded data can be uploaded to the server device and then listened to by the client device. Further, in the audio distribution system of Patent Document 2, when streaming audio data from a server apparatus to a plurality of client apparatuses, the reading speed is controlled so that the audio data buffer amount can maintain a target value, and reproduction of all client apparatuses is performed. Can be synchronized. Furthermore, there is a net radio system that uses a portable telephone, an Internet telephone service called “Skype”, and a personal computer (PC) to deliver a program mainly in audio data in real time on the Internet.

JP 2005-316891 A JP 2005-148565 A

  Here, when it is considered that real-time data such as voice data is distributed from the source device that is the source of information to the client device via the server device, information is transmitted while the communication line between the source device and the server device is interrupted. It is not possible to send itself. For example, if the source device is a mobile phone, communication with the server device will be interrupted if it reaches a place where it is difficult to receive radio waves in mountainous areas on a train, and when communication resumes later, It was amazing. It ’s a shame that I could n’t tell you live. ”

  In view of such circumstances, the present invention provides a method in which real-time data such as audio data is distributed from the source device to the client device via the server device even if the communication line between the source device and the server device is interrupted. It is an object of the present invention to provide a data distribution system that can prevent data distributed to a device from being interrupted.

  According to the main feature of the present invention, a source device (RC: for example, RCa), a server device (SV), and one or more client devices (CL: for example, CLa1, CLa2), a real-time data source that is a data supply source A data distribution system for distributing data from a device (RC) to a client device (CL) via a server device (SV), wherein the source device (RC) buffers source buffering means (2B; R3). Can communicate with the server device (SV) by the source communication state detection means (R6, R7) for detecting the communication state between the server device (SV) and the source communication state detection means (R6, R7). While being detected (R4 = YES), the data buffered by the source buffering means The first transfer means (R5) that sequentially transfers to the server apparatus (SV) at the time rate and the communication with the server apparatus (SV) become impossible again by the source communication state detection means (R6, R7). When it is detected (R6 → R4 = NO → R8 = YES), the data is buffered by the source buffering means (2B; R3) while communication is disabled prior to the data transfer by the first transfer means (R5). Second transfer means (R9) for transferring the ringed data to the server apparatus (SV) at a rate faster than the real time rate. The server apparatus (SV) includes the first transfer means (R5) and the second transfer. Server buffering means (SB: S7, S8) for buffering data (Drs, Drb) transferred by the means (R9), and server buffering means (SB: S7, S8) ) Buffered data distribution means (S11) for delivering the data (Ds) buffered by the client device (CL) at a real time rate, data (Drb) transferred by the second transfer means (R9), and this Speed conversion means (SA: S8) for shortening the time axis of the data (Drs) transferred by the first transfer means (R5) for a predetermined time following the data transfer from the real time, and server buffering means ( SB: S7, S8) provides a data distribution system (claim 1) for buffering data whose time axis is shortened by the speed conversion means (SA: S8) (SB: S8). Note that the parentheses are reference symbols, terms and the like of the examples added for convenience of understanding, and the same applies to the following.

  A data distribution system according to the present invention comprises (a) a source device (RC: for example, RCa), a server device (SV), and one or more client devices (CL: for example, CLa1, CLa2). Distribution is performed from the source device (RC), which is a supply source, to the client device via the server device (SV). The source device (RC) buffers real-time data such as voice, music, and video (2B; R3), and in a normal state where communication is possible, the buffered data (Drs) is sent to the server device (SV) at a real-time rate. (R5), when it is detected that communication with the server apparatus (SV) is disabled again (R6 → R4 = NO → R8 = YES), communication is not possible. The buffered data (Drb) is transferred to the server device (SV) at a rate faster than the real time rate (R9). The server device (SV) is the normal time (S5 = YES → S6 = NO) data (Drs) transferred from the source device (RC) at the real-time rate, and the communication is restored at a rate faster than the real time ( The data (Drb) of S4 = YES) is buffered (S7, S8), and the buffered data (Ds) is distributed to the client device (CL) at a real time rate (S11).

  Therefore, according to the present invention, when real-time data such as audio data is distributed from the source device to the client device via the server device, even if the communication line between the source device and the server device is interrupted, it is distributed from the source device to the client device. The data being processed can be kept from being interrupted.

  In the data distribution system according to the present invention, when the communication between the source device (RC) and the server device (SV) is resumed in the server device (SV) (S4 = YES), the real time rate is transmitted from the source device (RC). Data transferred at a faster rate (data Drb buffered by the source device (RC) during communication failure) and a predetermined time following this data transfer [buffering data Db (Pr to Pw) is the minimum time Minutes (until S6 = YES)] Speed conversion is performed to shorten the time axis of the data (Drs) transferred at the real time rate (SA: S8), and the data with the time axis shortened (SA: S8). Da) is buffered (SB: S8).

Therefore, according to the present invention, when the communication between the source device and the server device is resumed even though the data is distributed to the client device at the real time rate, the client device reduces the time axis by the server device. Based on the data that has been speed-converted, voice, music, video, etc. are played back at high speed, the delay when communication is restored is restored, and the original data reception state with the minimum delay is restored. can do.

1 shows a configuration example of a data distribution system according to an embodiment of the present invention. 4 shows an example of a source device data transfer processing flow according to an embodiment of the present invention. The example of the server apparatus data delivery processing flow by one Example of this invention is shown. It is buffering explanatory drawing in the server apparatus by one Example of this invention. It is operation | movement explanatory drawing of the client apparatus by one Example of this invention.

[System configuration example]
FIG. 1 shows a system configuration example according to an embodiment of the present invention, FIG. 1 (1) is a hardware configuration example of a source device, and FIG. 1 (2) is an overall configuration example of a data distribution system. As shown in FIG. 1 (2), the entire data distribution system includes source devices RC: RCa; RCb;..., Server devices SV and client devices CL: CLa1, CLa2,. The reference symbols “RC” and “CL” represent the source device and the client device, respectively. As the source device RC, a portable communication terminal device is used like a mobile computer with a voice input function equipped with a mobile phone or wireless communication means. In the example of FIG. 1 (1), a central processing unit (CPU) 1, random Access memory (RAM) 2, read-only memory (ROM) 3, external storage device 4, setting operation input device 5, display device 6, A / D conversion circuit 7, D / A conversion circuit 8, communication interface (communication I / O F) Elements 9 and the like are provided, and these elements 1 to 9 are connected to each other via a bus 10.

  The CPU 1 that controls the entire apparatus, together with the RAM 2 and the ROM 3, constitutes a data processing unit that executes various processes according to various control programs, and includes various types of source apparatus data transmission processing using a clock by the timer 11 according to the various control programs. Execute the process. The RAM 2 temporarily stores various data necessary for these processes, and includes a buffer memory 2B that sequentially holds and sequentially outputs audio data to be transmitted to the server device SV in addition to a normal work area for this purpose. Yes. The ROM 3 can store various control programs including a source data transmission program and necessary control data.

  The external storage device 4 is a hard disk (HD), a compact disk read only memory (CD-ROM), a flexible disk (FD), a magneto-optical (MO) disk, a digital multipurpose disk (DVD), a memory card, etc. The storage means uses a storage medium, and can record any necessary data.

  The setting operation input device (input operation unit) 5 has a panel operation unit for performing various setting operations, and introduces the contents of the panel operation by the user at the panel operation unit to the data processing unit as setting operation information. The display device (display unit) 6 controls display contents of a display (display device such as an LCD) connected to the display device 6 according to a command from the data processing unit, provides display assistance for the operation of the setting operation input device 5, and The video based on the video data can also be displayed.

  The A / D conversion circuit 7 is connected to a microphone device 12 that converts a voice input by a source device user into an input voice signal. The A / D conversion circuit 7 performs A / D conversion on the input voice signal from the microphone device 12 and converts the input voice data into data. Installed in the processing unit (buffer memory 2B). The D / A conversion circuit 8 is processed by the data processing unit (2B) and transmitted to the outside of the source device, external voice data sent from the outside of the source device, or guidance voice data generated by the data processing unit. Is D / A converted, and an output audio signal is sent to the speaker device 13, and the output audio is emitted from the speaker device 13. The D / A conversion circuit 8 and the speaker device 13 are not essential components of the source device RC and can be omitted.

  The communication I / F 9 is various interfaces that are wirelessly connected to a communication network CN such as the Internet and a telephone line, and can exchange various information with the server device SV and other communication terminals connected to the communication network CN. . When the data distribution system is formed, the server apparatus SV is connected to the communication network CN as shown in FIG. 1 (2), and voice data is transmitted from the source apparatus RC to the server apparatus SV.

  The hardware of the server device SV is configured in the same manner as the source device RC of FIG. 1A, but the A / D conversion circuit 7, the D / A conversion circuit 8, the microphone device 12, and the speaker device 13 are not necessary. In the typical server device SV, the communication I / F is connected to the communication network CN by wire. In addition, audio data transmitted from the source device RC is introduced into the buffer memory (SB) provided in the RAM via the communication network CN and the communication I / F, and the audio data output from the buffer memory (2B) is And transmitted to the client device CL via the communication network CN and the communication I / F.

  The hardware of the client device CL is also configured in the same manner as the source device RC of FIG. 1 (1), but the A / D conversion circuit 7 and the microphone device 12 are unnecessary, and it is not necessary to provide a buffer memory. The connection between / F and the communication network CN may be wireless or wired. When the data distribution system is formed, the communication network CN is connected not only to the source device RC but also to the client device CL as shown in FIG. 1 (2), and the voice data from the source device RC is transmitted to the server device. It is transmitted from the SV to the client device CL. In this case, in the client device CL, the audio data input from the communication I / F is D / A converted by the D / A conversion circuit, and the corresponding audio is emitted from the speaker device.

[Basic operation of data distribution system]
One data distribution system that distributes one data includes one source device RC, one server device SV, and one to a plurality of client devices CL. The example of FIG. 1B is an example of a source device A (RCa). -A common server device (SV)-Two client devices A1, A2 (CLa1, CLa2) form a first data distribution system, and source device B (RCb)-Common server device (SV)-Two It is shown that the second data distribution system is formed by the client apparatuses B1 and B2 (CLb1 and CLb2).

  Taking the first data distribution system as an example, the basic configuration of a data distribution system according to an embodiment of the present invention will be described. The data distribution system includes a source device RCa, a server device SV, and one or more client devices CLa1. , CLa2 and distributes real-time data such as audio data from the source device to the client device via the server device. The source device RCa sequentially inputs and buffers real-time data input from the voice input means 12-7 or the like at the real time rate in the buffer memory 2B. In a normal state in which communication with the server device SV is possible, the source device RCa uses the buffer memory 2B. The buffered data is read from the top at the real time rate and transferred to the server device SV. Here, when communication with the server apparatus SV is disabled, reading from the buffer memory 2B is stopped, real-time data is accumulated in the buffer memory 2B, and communication is restored to the original normal state in which communication with the server apparatus SV is possible. The buffered data stored in the buffer memory 2B during the disabled state is actually read from the portion immediately after the stop of reading (the data portion immediately after the data portion last sent to the server device SV immediately before communication is disabled). Read at a rate faster than the time rate and transfer to the server device SV. The server device SV sequentially inputs data transferred from the source device RCa at a real time rate in a normal state and data transferred at a rate faster than the real time rate when communication is restored to the buffer memory of the server device SV. And deliver the buffered data to the client devices CLa1 and CLa2 at a real time rate.

  Note that the server device SV can distribute other data prepared in advance to the client devices CLa1 and CLa2 while communication with the source device RCa is not possible. Also, conversion (speed) that shortens the time axis of data buffered by the source device RCa while communication is disabled and transferred at a high rate when communication between the source device RCa and the server device SV is resumed. Conversion), and buffer the data that has been speed converted to shorten the time axis. That is, although data is delivered from the server device SV to the client device CL at a real time rate, when the communication between the source device RCa and the server device SV is resumed, the client device CL is based on the speed-converted data. The voice is reproduced at a high speed, the delay when the communication failure is restored is recovered, and when the minimum delay is reached, the original data reception state is restored.

[Operation of each device]
Hereinafter, as shown in FIG. 1 (1), the operation of each device when audio data is used as input real-time data will be described. 2, 3, 4, and 5 are diagrams for explaining operations of the source device, the server device, and the client device in the data distribution system according to the embodiment of the present invention. 5 (1) shows a flowchart representing an example of source device data transfer processing, server device data distribution processing, and client device data reception processing, respectively. In the source device RC, when the source device data transfer process is started by a predetermined operation and there is an operation for accessing the server device SV and instructing the publication of the program (execution of the data distribution service), the CPU 1 performs step R1 in FIG. Then, the source device RC is connected to the server device SV to instruct the release of the program. Thereby, operation | movement of the data delivery system which delivers the audio | voice data from the source device RC starts.

  On the other hand, in the server device SV, the CPU accepts the connection from the source device RC in step S1 of FIG. 3, and can distribute the distribution service of the audio data transferred from the source device RC according to the program release instruction. The program display data is created in accordance with the program information of the source device RC registered in advance in the server device SV, the audio buffer for the source device RC (FIG. 4: SB) is cleared, and the source device RC Audio data is received, buffered, and prepared for distribution to the client device CL.

  In the source device RC, after connecting to the server device SV in step R1 and instructing program release, in step R2, the buffer memory 2B is cleared and preparation for buffering of audio data input from the audio input means 12-7 is made. do. When the audio data is input, the process proceeds to step R3, where the input audio data is sequentially acquired and loaded in the buffer memory 2B. That is, when the program release is instructed, the audio input from the microphone 12 is A / D converted by the A / D conversion circuit 7 and stored in the buffer memory 2B of the RAM (main storage device) 2. The buffer memory 2B is also called an audio buffer, and sequentially writes input audio data at a position indicated by a recording pointer (also called a write pointer) at a real time rate. The recording pointer is incremented as time elapses regardless of whether communication with the server device SV is possible, and the writing position of the audio data is sequentially advanced.

  In the next step R4, it is determined whether or not the source status flag is “normal”. Here, the source status flag indicates whether communication between the source device RC and the server device SV is possible according to the detection output of the communication status detection means provided for detecting the communication status with the server device SV. This is a flag indicating whether or not communication is possible. If communication with the server SV is possible, the source status flag is “normal”, and the read pointer of the audio buffer 2B is incremented as time elapses (recording pointer is incremented). Then, the voice data reading position is sequentially advanced. When the source status flag = “normal”, for example, audio data for one stream transfer is written from the read position indicated by the read pointer to immediately before the write position indicated by the recording pointer. On the other hand, when communication with the server SV is disabled, the source status flag = “interrupt”, and the increment of the read pointer is stopped, and the audio data reading is interrupted. As a result, while the communication between the source device RC and the server device SV is interrupted, the audio data is recorded in the audio buffer 2B.

  When it is determined in step R4 that the source state flag is “normal” (R4 = YES), the process proceeds to step R5, and the audio data Drs in the buffer 2B is read at the real time rate according to the instruction of the read pointer, and communication is performed. Stream transfer (also referred to as streaming transfer) is performed wirelessly from the I / F 9 to the server device SV via the communication network CN. In the next step R6, the communication state with the server apparatus SV is checked. If it is bad, the source state flag is changed to “interrupted”, and if it is good, “normal” is maintained. In step R7, the read pointer of the audio buffer 2B is updated, and the process returns to step R3.

  On the other hand, when it is determined that the source status flag is not “normal” but “suspended” (R4 = NO), the process proceeds to step R8, where it is determined whether the communication status with the server device SV is good, If it is still defective (R8 = NO), the process returns to step R3. When it is determined in step R8 that the communication state with the server apparatus SV is good, that is, when the communication state with the server apparatus SV is restored (R8 = YES), the process proceeds to step R9, and the voice buffer 2B is started from the point of interruption start. The voice data DRb stored in is sent to the server SV at a stretch faster than the real time rate by bulk transfer, and the source state flag is set to “normal” in the next step R10. In step R7, the read pointer of the audio buffer 2B is updated, and the process returns to step R3.

  After returning to step R3, the processing of steps R3 to R10 described above is repeated according to the progress of voice input and the communication state, and there is an operation for instructing the end of the program release, and this instruction is confirmed by the server device SV. (Not shown), the source device data transfer process is terminated.

  In the client device CL, when the client device data reception process is started by a predetermined operation and there is an operation for accessing the server device SV and requesting a list of programs that can be distributed, the CPU performs step C1 in FIG. The client device CL is connected to the server device SV to request a distributable program list. Thereby, the client device CL can participate in this data distribution system.

  In the server apparatus SV, as described above, after the CPU receives the connection from the source apparatus RC and prepares the voice data distribution from the source apparatus RC in step S1 of FIG. 3, the list request from the client apparatus CL is issued. When received, in the next step S2, a program list is returned to the client device CL in response to a request from the client device CL. This program list is a list of programs (audio data distribution service) that can be distributed from the server apparatus SV to any client apparatus CL based on the program display data created in accordance with the program release instruction from each source apparatus RC. And includes necessary reference information such as server state information (Cn) indicating the current communication state between the server device SV and each source device RC.

  Accordingly, when the client device CL acquires the distributable program list from the server device SV in step C1 of FIG. 5 (1), the acquired distributable program list is presented as a program selection screen on the display of the display device. Let the device user select a desired program. FIG. 5 (2) shows an example of a program selection screen. In this example, in addition to basic information such as the name of the currently released program and personality (nickname of the source device user), the program is released by each source device RC. For example, “elapsed time” expressed in “hour: minute”, server state information Cn described above, “sound quality” of voice data transferred from each source device RC, “listener's” Real-time reference information such as “number of people” (the number of client devices CL currently receiving the program) is included in the display items. The server device SV is provided with communication state monitoring means for detecting and monitoring the communication state between the server device SV and the source device RC, and the server state information depends on the communication state detected by the communication state monitoring means. The communication state between the server device SV and each source device RC is represented by “state” such as “Good” (good, normal), “disConnected” (interrupted). Therefore, when the client device user performs a program selection operation on the displayed program selection screen, the server device SV is requested to distribute the selected program in accordance with the program selection operation in the next step C3.

  On the other hand, in step S3 in FIG. 3, the server apparatus SV adds the client apparatus CL to the distribution destination (listener) of the program requested for distribution in response to the program distribution request from the client apparatus CL. Proceeding to each program distribution processing stage SP, distribution processing is performed for the requested program. As shown in FIG. 3, each program distribution processing stage SP includes the first half of the buffering processing portion SP1 for loading audio data from the source device RC into the buffer memory (SB) = steps S4 to S8 and the buffer memory (SB). The latter half stream distribution processing part SP2 for streaming the audio data to the client device CL = steps S9 to S12.

  In the program distribution processing stage SP, first, the reception state of the audio data is checked, and the process proceeds to the first step S4 to start the buffering process SP1. In step S4, it is determined whether or not audio data has arrived by bulk transfer. If there is no audio data bulk transfer (S4 = NO), the process proceeds to step S5 to determine whether or not audio data Drs has arrived by stream transfer. judge. Here, when the audio data Drs is stream-transferred (S5 = YES), the process further proceeds to step S6, where the audio data (Db) stored in the buffer memory (SB) is examined, and the already received audio data It is determined whether or not the data is accumulated in the buffer memory (SB). If the data is not accumulated (S6 = NO), the audio data received by the stream transfer is loaded in the buffer memory (SB) as it is in step S7.

  On the other hand, when it is determined in step S4 that the audio data Drb has arrived by bulk transfer (S4 = YES), or when the audio data Drs has arrived by stream transfer in step S6, the received data is stored in the buffer memory (SB). When it is determined that it has accumulated (S6 = YES), in step S8, the audio data Drb and Drs arrived by bulk transfer or stream transfer are speed-converted and loaded in the buffer memory (SB).

  FIG. 4 is a diagram for explaining buffering in the server apparatus according to one embodiment of the present invention. Here, the buffering function of the server apparatus will be described with reference to FIG. The server device SV buffers the audio data Drs and Drb from the source device RC, and in order to stream the buffered audio data Ds to the client device CL, a speed control unit (also referred to as a time control unit) SA, The data processing unit includes the functions of the buffer memory SB, the bulk transfer detection unit SC, and the accumulation detection unit SD.

  The speed control unit SA normally operates so that the audio data Drs stream-transferred from the source device RC at a real time rate is input to the buffer memory SB as it is. However, when the communication is resumed after the communication between the server device SV and the source device RC is interrupted, the speed conversion function is activated by detecting the bulk transfer of the voice data by the bulk transfer detection unit SC, and the recovery from the source device RC is started. Then, speed conversion that shortens the time axis of the audio data Drb that has been bulk transferred at a high rate and the audio data Drs that is subsequently stream transferred at the real time rate, that is, so-called speech speed conversion or fast rotation, is performed. Then, the voice data Da that has been speed-converted with the time axis shortened by the speed conversion function of the speed controller SA is input to the buffer memory SB. The accumulation detection unit SD examines the audio data Db accumulated in the buffer memory SB. If the audio data Db exceeding a predetermined minimum time (one stream delivery to the client device CL) is accumulated, Is determined. The speed conversion function of the speed control unit SA is executed while the audio data Db is accumulated in the buffer memory SB, and the audio data Db reaches a predetermined minimum time. If it is determined that it is gone, it stops. That is, the data Da obtained by converting the speed of the audio data Drb and Drs from the source device RC is input to the buffer memory SB until the normal operation is resumed after the communication is recovered (the operation state during this period is called the chasing state). .

  The buffer memory SB is provided in the RAM (main storage device) of the server device SV and is also called a data buffer, and writes the audio data Drs and Da input from the speed control unit SA at a position indicated by the write pointer Pw. When writing is performed, the write pointer Pw is incremented by the written time. Therefore, while the communication with the source device RC is interrupted (interrupted), the audio data Drs and Drb are not sent, so the write pointer is not updated. On the other hand, while the communication with the source device RC is normal with respect to the client device CL, the playback pointer (also referred to as a read pointer) Pr is incremented as time elapses, and the audio data Ds from the position indicated by the playback pointer Pr. While the communication with the source device RC is interrupted, the playback pointer Pr is not incremented and the audio data Ds is not read from the data buffer SB.

  Now, when it is determined in step S5 that the audio data is not stream-transferred (S5 = NO), in step S7, the process of loading the audio data Drs delivered by the stream transfer as it is in the data buffer SB is finished. Alternatively, in step S8, the audio data Drb and Drs delivered by bulk transfer or stream transfer are converted in speed according to the speed conversion function of the speed control unit SA, and the process of loading the speed-converted audio data Da in the data buffer SB is completed. If this is the case, the buffering process SP1 is once ended, and the process proceeds to step S9.

  In step S9, it is determined whether or not the timing of the stream delivery cycle has been reached. If the stream delivery timing has not been reached (S9 = NO), the process returns to the first step S4 of the buffering process SP1, and the above-described step S4 The buffering process SP1 of .about.S8 is repeated. On the other hand, when the stream distribution timing is reached (S9 = YES), the process proceeds to step S10, and the stream distribution process SP2 is started.

  In step S10, it is determined whether or not the communication state with the source device RC is good by the communication state monitoring means, and if good (S10 = YES), the process proceeds to step S11 and the reproduction pointer in the data buffer SB is reached. The audio data Ds at the position indicated by Pr is read out and streamed to the client device CL, and then the process returns to step S4. If the communication with the source device RC is interrupted (S10 = NO), the process proceeds to step S12, and the audio data De of another program based on the content prepared in advance in the server device SV is stream-distributed to the client device CL. Then, the process returns to step S4. Here, as the audio data De of another program, for example, when message content or music content is recorded in advance for communication interruption and communication with the source device RC is interrupted, “Now the communication line is bad and cannot be relayed. “Please wait for a while” or try to play a message repeatedly, or “Please listen to music for a while.”

  After returning to step S4, the processing of steps S4 to S12 described above is repeated according to the data transfer state from the source device RC, the data accumulation state of the data buffer SB, and the communication state with the source device RC. If there is an instruction to end the program release, the client device CL is notified of the end of the program (not shown), and the program distribution processing SP related to the program is ended.

  On the other hand, in the client device CL, based on the audio data Ds stream-distributed via the server device SV using the source device RC as the transmission source in step C4 of FIG. 5 or based on the content prepared in the server device SV. The audio data De of another program that is stream-distributed from the server SV itself is received. In the next step C5, the received audio data Ds and De are reproduced by a D / A conversion circuit and a speaker. Thus, the processing of steps C4 and C5 is repeated until the end of the program is notified from the server device SV, and the user of the client device CL enjoys the program selected by the client device CL. Then, upon receiving a program end notification from the server device SV, the client device data receiving process is terminated.

[Various Embodiments]
Although a preferred embodiment of the present invention has been described above, this is merely an example, and the present invention can be variously modified and implemented in various modes without departing from the spirit of the invention. Can do. For example, stream transfer to a server device or stream delivery to a client device is performed using a general method. The sender sends audio data to the receiver every short period (for example, 10 msec) or every predetermined number of samples (4800 samples). At that time, the sender side may send it by a protocol such as TCP, and it may be confirmed every time whether it is received by the receiver. Alternatively, as in the RTP protocol, the sender side may send by a protocol such as UDP that cannot confirm the reception of the other party, and in parallel may be confirmed by the receiver by TCP or the like.

  Alternatively, the receiver may make a request to the sender side with a technique such as TCP. The recipient can know the communication failure when the request is not received.

  The detection of the communication state may be inferred from the radio wave state. Alternatively, the transmission side may try to communicate once and know that it has failed. Alternatively, the receiving side may know that data that should have arrived at a predetermined timing does not arrive.

  The conditions for streaming data in the data buffer (SB) from the server device to the client device are good communication with the source device (S10 = YES) in the embodiment, but the data stored in the data buffer (SB) It may be set when there is no (Db) or less than a predetermined time (pointers Pr and Pw match or are adjacent).

  In the embodiment, the speed conversion (time axis shortening) is performed to gradually recover the delay when recovering from the communication failure, and eventually the delay between the source device and the client device is minimized. The speed conversion (time axis shortening) may not be performed. When speed conversion is not performed, the delay when recovering from the failure remains as it is, but a delay corresponding to the size of the data buffer (SB) provided in the server device is allowed.

  In the embodiment, the speed conversion is performed by the server device. However, the speed conversion may be performed when the source device performs bulk transfer.

  A general method for converting the speed of audio data is used. In other words, voice data is spliced in the time direction. For example, the volume is crossfade through a window according to the volume change shape, such as speech speed conversion, and the time axis is shortened by considering the silent part. To do.

  The data to be distributed is not limited to sound but may be video and performance data (MIDI). In the case of video, drop frames to increase the speed from real time. In the case of performance data (MIDI), the event is recorded and buffered with the delta time, and when the speed is made faster than the actual time, the tempo that is the basis of the delta time is advanced from that at the time of recording.

RC: RCa, RCb, ... source device,
SV server device,
CL: CLa1, CLa2,...; CLb1, CLb2,... Client device,
Drs, Drb Stream transfer and bulk transfer audio data,
Da, Db Speed-converted audio data and buffered audio data,
Ds, De Audio data from the streamed buffer and audio data of another program.

Claims (1)

A data distribution system that includes a source device, a server device, and one or more client devices, and distributes real-time data from the source device, which is a data supply source, to the client device via the server device,
The source device is
Source buffering means for buffering real-time data;
Source communication state detection means for detecting a communication state with the server device;
First transfer means for sequentially transferring the data buffered by the source buffering means to the server apparatus at a real time rate while it is detected by the source communication state detection means that communication with the server apparatus is possible;
When it is detected by the source communication state detection means that communication with the server device is disabled, the source buffering means during the communication disable prior to the data transfer by the first transfer means. Second buffer means for transferring the buffered data to the server device at a rate faster than the real-time rate,
Server device
Server buffering means for buffering data transferred by the first transfer means and the second transfer means;
Buffering data distribution means for distributing data buffered by the server buffering means to the client device at a real time rate;
Speed conversion means for shortening the time axis of the data transferred by the second transfer means and the data transferred by the first transfer means for a predetermined time following the data transfer from the actual time,
The server buffering means buffers data whose time axis has been shortened by the speed converting means.

Images