JP5524374B2 - Video stream transmission method - Google Patents

Description

  The present invention relates to a video server apparatus that supplies a video stream to a client device via a network.

  With the spread of video sources using video compression technology such as MPEG (Moving Picture Experts Group) -2, video signals have been transmitted over computer networks such as Ethernet (registered trademark). However, in IP packet transmission over Ethernet (registered trademark) and the like, there is no guarantee of quality of service (QoS (Quality of Service)), especially in the Internet and wireless LAN, etc., which is sufficient to support the bit rate of the original video stream. There is often a problem that the transmission band cannot be secured, and in that case, there is a problem that the video cannot be seen at all.

  As an improvement measure for this problem, when the receiving apparatus estimates the information transmission delay using the reception bit rate and the information encoding speed on the information transmitting apparatus side, and the estimated transmission delay is outside the target range There is an improved technique for maintaining transmission by calculating an appropriate encoding rate and requesting the information transmitting apparatus to change the encoding rate. As the above technique, for example, there is a technique (communication speed control method) disclosed in Patent Document 1.

  In addition, the network congestion state is determined by, for example, monitoring the reception status of video information, and the transmission of video information on the transmission side is controlled based on the determination result. There are improved technologies that are trying to enable video streaming. As the above technique, for example, there is a technique (video information transmission method) disclosed in Patent Document 2.

JP-A-9-321798 (FIG. 1) JP2003-244695A (FIG. 1)

  With the above improvement technology, even when the network transmittable bandwidth is lower than the bit rate of the original video stream, the stream that can be decoded and displayed by the receiving device is reduced by reducing the bit rate of the stream sent to the network. It becomes possible to deliver via. However, the improvement technique basically requires processing such as returning the result of monitoring the stream status on the receiving device side to the server device side, and the receiving device uses a device suitable for the server device. There was a problem of putting a burden on the side.

The present invention has been made in order to solve the above-described problems. Even if the transmission bandwidth of the network is lower than the bit rate of the original video stream without imposing a burden on the receiving device side, it is continuously in real time. It is an object of the present invention to obtain a video stream transmission method capable of continuing streaming.

The video stream transmission method according to claim 1 of the present invention performs HTTP communication between a video server device and a receiving device connected via a network, establishes communication using TCP / IP, and transmits data. A video stream transmission method for reading the video stream from the video server device, converting a bit rate related to the read video stream into a set conversion bit rate, and converting the video into the set conversion bit rate Including a method of generating a transmission packet including position information of a stream to be transmitted to a network and transmitting the transmission packet to the reception device, and the setting conversion bit rate is set for a plurality of the transmission packets. The location information contained in each and the sending packets are sent out. If the average transmission bit rate obtained based on a time difference which is below the set transformation bit rate is the average controlled to be less than the transmission bit rate.

In the video stream transmission method according to the first aspect of the present invention , real-time streaming can be continuously continued even when the transmittable bandwidth of the network is lower than the bit rate of the original video stream.

It is a block diagram which shows the structure of the video server apparatus which is Embodiment 1 of this invention. It is a block diagram which shows the structure of the video server apparatus which is Embodiment 2 of this invention.

<Principle of the invention>
Stipulates a process for recognizing the existence of a server on the network from the client device and selecting the video content to be acquired when supplying a video stream from the video server device to a TV receiver or the like as a client via the network There is a need to. Standardization activities such as Universal Plug and Play (UPnP) and Digital Living Network Alliance (DLNA) have been promoted in order to realize interoperability between PCs (personal computers) and home appliances regardless of manufacturer. Yes. In DLNA, the server device that supplies the video stream is named Digital Media Server (DMS), the client device that receives and decodes and displays the stream is called Digital Media Player (DMP), and HTTP (HyperText Transport Protocol) is used as the baseline for stream transmission. It is prescribed to use.

  As long as the video server device of the present invention is a receiving device that complies with the standard that stipulates that HTTP is used as the baseline for stream transmission, the network server can be used without adding or modifying equipment other than the video server, such as a receiving device. The stream can be supplied at a bit rate that matches the transmittable band.

  Specifically, the video server device of the present invention wave inserts the position information from the head of the stream into the MPEG-2 transport stream to be sent to the network, and then sends the position information to the network as a TCP / IP packet. The bit rate of the stream that can be transmitted on the network is estimated from the time when the included packet is transmitted and the position information. And by converting the bit rate of the video stream based on the estimated transmittable bit rate and sending it out, it is possible to supply a receivable stream even to a receiving device that cannot transmit using the original stream before conversion. It is said.

<Embodiment 1>
FIG. 1 is a block diagram showing a configuration of a video server apparatus according to Embodiment 1 of the present invention. As shown in the figure, a hard disk drive (HDD) 1 as storage means stores a read stream to be transmitted. The read control means 2 reads the read stream stored in the HDD 1, extracts a video (video) stream (video signal data string), and supplies it to the transcoder 3 (transcode means). At this time, the read control means 2 reads at an appropriate speed based on the time information included in the read stream.

  The transcoder 3 converts the bit rate of the video stream at the set conversion bit rate indicated by the conversion control means 7. The stream generation means 4 generates a transmission packet (transmission stream) from the stream obtained via the transcoder 3.

  The HTTP server 5 establishes TCP / IP communication with the client device via Ethernet (registered trademark), issues a stream read start instruction to the read control means 2 based on the stream transmission request by HTTP, and the stream generation means 4. The transmission packet generated in step (1) is sent to a network interface (I / F) 6 for Ethernet (registered trademark). By outputting a transmission packet through the network I / F 6, it is possible to connect to an Ethernet (not shown). As a result, streaming transmission can be performed by transmission packets to various receiving devices (client devices) connected to the network I / F 6.

  The sending packet monitoring means 8 monitors packets sent from the network I / F 6. The conversion control means 7 controls the output bit rate of the transcoder 3 based on the monitor information S8 obtained from the sending packet monitor means 8.

  The operation of the video server device according to the first embodiment will be described below. Here, it is assumed that the transmitted stream takes the form of a transport stream defined in the MPEG-2 system standard. Further, it is assumed that this video server apparatus supplies a stream using a TCP / IP protocol to a client device connected via Ethernet (registered trademark).

  In addition, here is a client that supports MPEG-2 transport stream transmission over HTTP, using devices designed based on the standards such as Universal Plug and Play (UPnP) and Digital Living Network Alliance (DLNA). It is assumed that a device (such as a digital media player (DMP)) is connected.

  Streaming transmitted from the video server device according to the first embodiment is started when a client device connected to Ethernet (registered trademark) establishes a TCP / IP connection to the HTTP server 5 and sends an HTTP-GET command. . The HTTP server 5 instructs the read control means 2 to read out the video content file specified by HTTP-GET from the HDD 1. The read control means 2 reads the designated file (MPEG-2 stream, read stream) from the HDD 1. The MPEG-2 stream read by the read control unit 2 is sent to the HTTP server 5 via the transcoder 3 and the stream generation unit 4.

  In the transcoder 3, the MPEG-2 stream (program stream or transport stream) read by the read control means 2 is first separated into video and audio streams, and then the video stream is designated by the conversion control means 7. The setting conversion bit rate is converted. Both the converted video stream (video stream after bit rate conversion) and the original audio stream (through the transcoder 3 without being bit rate converted) are sent to the stream generation means 4.

  Here, the transcoding function of the transcoder 3 has a function of reducing the bit rate of the input video stream based on the set conversion bit rate. To realize this function, a dedicated integrated circuit is used, or an MPEG-2 video decoder and an MPEG-2 video encoder are combined and implemented by re-encoding the decoded video stream at a desired bit rate. Can be considered.

  The stream generation means 4 multiplexes the audio stream sent from the transcoder 3 and the converted video stream subjected to the bit rate conversion, and collects them into an MPEG-2 transport stream (transmission packet). The MPEG-2 transport stream is generated by dividing and inserting video data, audio data, system information (PAT, PMT) indicating the stream configuration, etc. into a payload of a packet with a header of 188 bytes. An ID number (PID) is attached to the header of each packet, and the type of data of each packet is specified by the PID. The association between the data type and the PID is performed using a program association table (PAT) and a program map table (PMT). PAT has PID fixed to “0” and includes PID information of PMT that holds mapping information to PID of data included in each program for all programs included in the transport stream.

  The stream generation means 4 converts the above tables, video, and audio data into TS packets based on the PID determined in accordance with the MPEG-2 system standard in advance, and sets the number of packets from the beginning in the payload for each predetermined period. The inserted transmission rate measurement packet is inserted into the transmission packet.

  As the PID of the transmission rate measurement packet, a value that does not overlap with the PID assigned for a specific purpose in the MPEG-2 system standard or the PID of a packet including actual data such as video and audio is selected. The transmission packet including the transmission rate measurement packet generated by the stream generation means 4 is supplied to the HTTP server 5 and transmitted to the network via the network I / F 6.

  On the other hand, the IP packet (outgoing packet) sent out from the network I / F 6 is read by the sending packet monitor 8 at the time of sending out the network, and only the transmission rate measurement packet inserted by the stream generating means 4 is filtered by the PID and data transmission is started. The number of packets sent out from is read out and stored together with the time when the transmission rate measurement packet is read.

  The sending packet monitoring means 8 estimates the network transmittable bandwidth from these values (reading time, number of packets in between). As an example, when the transmission rate measurement packet that is nth from the beginning is read at time Tn, and the transmission rate measurement packet packet that is mth from the beginning is read at time Tm, the following equation (1) holds: To do.

Br = (mn) .188 / (Tm-Tn) (1)
The sending packet monitoring means 8 monitors the average bit rate Br calculated by the above equation (1), assumes that a stream is flowing, and sends monitor information S8 defining the average bit rate Br to the conversion control means 7. Tell. That is, the sending packet monitoring means 8 can obtain the average bit rate Br which is the bit rate of the stream that can be transmitted, based on the sending time of the transmission rate measuring packet.

  By the way, when such stream transmission is performed, a buffer for temporarily storing a reception stream is generally provided on the reception side before the decoding process. For this reason, when the transmission band is secured to some extent, if the amount of data stored in the reception buffer is small, the stream is sent out at a bit rate higher than the bit rate of the transmitted stream, or conversely, it is already in the reception buffer. If sufficient data is accumulated, the amount of data to be transmitted may be reduced. However, the average bit rate Br with high accuracy can be obtained by observing the bit rate for a sufficiently long period, for example, (Tm−Tn) being 1 second or longer.

  At this time, the MPEG-2 decoder used on the receiving side is synchronized with the time stamp (program clock reference: PCR) defined by the MPEG-2 system standard of the transport stream that is generally sent or received. It operates according to the internal clock of the side and tries to acquire data at a rate equal to the bit rate of the stream. Therefore, the average bit rate obtained in this way is approximately equal to the bit rate of the stream if the transmission bandwidth is sufficiently wide.

  However, if the transmission bandwidth cannot be secured sufficiently and is lower than the bit rate of the stream, the average bit rate Br observed at the time interval including the period in which the receiver stores data in the buffer at the start of transmission is calculated by the video server device. The bit rate is lower than the bit rate of the stream that can be transmitted, and is an index indicating the maximum value of the bit rate that can be received on the receiving side.

  Therefore, in order to continue transmitting the stream so that the decoder on the receiving side can continue the decoding operation without failure, it is necessary to reduce the bit rate of the transmission stream below the average bit rate Br obtained by Equation (1).

  For this reason, the sending packet monitor 8 monitors the average bit rate Br calculated by the above equation (1) from the read packet, and transmits monitor information S8 indicating the average bit rate Br to the conversion control means 7.

  The conversion control means 7 controls the transcoder 3 so as to convert it into a stream having a transmittable bit rate based on the average bit rate Br indicated by the monitor information S8 received from the sending packet monitor means 8. That is, when the average bit rate Br monitored by the packet monitor unit 8 is smaller than the transmission bit rate Rb (set conversion bit rate), that is, when “Br <Rb”, the conversion control unit 7 sets the set conversion bit. The transcoder 3 is controlled so that the rate is equal to or less than the average bit rate Br.

  Note that when the video bit rate is converted by the transcoder 3, the image format may be the same as the input / output, or when the bit rate falls below a predetermined threshold, it may be changed to a low resolution screen.

  For example, if the video of the input stream to the transcoder 3 is HDTV and has a 1920 × 1080 pixel format and a bit rate of 20 Mbps, and the bit rate exceeds 10 Mbps, the HDTV image format remains as it is and 10 Mbps. When it falls below, it can be expected to effectively improve the image quality even at a low bit rate by changing to the SDTV format of 720 × 480 pixels.

  In this way, by providing the transcoder 3 with a resolution conversion function, by appropriately converting to an image format having a resolution suitable for the set conversion bit rate, a relatively good image quality suitable for the bit rate even at a low bit rate. There is an effect that can be maintained.

  Further, in the above-described bit rate determination method, there is a possibility that the set conversion bit rate may be changed in response to a temporary decrease in the average bit rate Br for some reason. Therefore, it is more stable by using it for controlling the set conversion bit rate based on the maximum value among the plurality of average bit rates Br indicated by the monitor information S8 obtained continuously from the packet monitoring means 8 at predetermined time intervals. Bit rate conversion operation can be performed.

<Embodiment 2>
FIG. 2 is a block diagram showing a configuration of a video server apparatus according to Embodiment 2 of the present invention. As shown in the figure, in addition to the configuration of the first embodiment shown in FIG. 1, a stream buffer 9 for temporarily storing the output stream of the stream generation unit 4 is provided between the stream generation unit 4 and the HTTP server 5. It has the structure inserted in.

  As described in Embodiment 1, when the bandwidth is sufficient, data is transmitted at a rate equal to the bit rate of the stream on average. Here, when the transcoder 3 is composed of a decoder / encoder combination, the hardware decoder / encoder is used so that the stream can be transmitted in real time, that is, at the transmission speed of the original MPEG-2 stream. When a transcoder is used, it is assumed that processing is performed at a speed equal to or faster than the original MPEG-2 stream transmission speed.

  For this reason, when reading the MPEG-2 stream stored in the HDD 1, the read control means 2 adjusts the read speed of the stream to match the stream bit rate based on time information such as a time stamp included in the stream. Thereby, transfer at the average data rate according to the bit rate of the stream is performed at the input / output of the transcoder 3. For this reason, if the bandwidth of transmission by Ethernet (registered trademark) is higher than the bit rate of the stream, the average bit rate of the stream written to the stream buffer 9 is equal to the rate read from here.

  If the transmission bandwidth is not sufficient and an extra delay occurs, the amount of data input to the stream buffer 9 becomes larger than the amount of output data, and the amount of data stored in the stream buffer 9 (storage) The data amount Bi) increases. Therefore, when the accumulated data amount Bi of the stream buffer 9 continues to increase, it can be determined that the bit rate of the stream to be transmitted needs to be reduced.

  Therefore, by sending the accumulated data information S9 defining the accumulated data amount Bi of the stream buffer 9 to the conversion control means 7, the conversion control means 7 can always recognize the change in the accumulated data amount Bi of the stream buffer 9.

  The conversion control means 7 recognizes the average bit rate Br defined by the monitor information S8 obtained from the sending packet monitoring means 8 in the same manner as in the first embodiment, and further stores the accumulation obtained from the stream buffer 9. The accumulated data amount Bi defined by the data information S9 can be recognized. Therefore, the conversion control means 7 performs transmission bit rate conversion control for the transcoder 3 by giving the set conversion bit rate to the transcoder 3 in combination with the average bit rate Br and the accumulated data amount Bi. Can do.

  For example, when the accumulated data amount Bi of the stream buffer 9 exceeds a predetermined threshold value Bth and the average bit rate Br calculated by the packet monitoring means 8 is smaller than the transmission bit rate Rb, that is, “ Only when Bi> Bth and Br <Rb, the control for executing the reduction control of the set anti-sense bit rate is possible.

  As described above, the video server apparatus according to the second embodiment controls the set conversion bit rate by the transcoder 3 based on the accumulated data information S9 from the stream buffer 9 in addition to the monitor information S8 from the transmission packet monitoring means 8. As a result, it is possible to control the set conversion bit rate with higher accuracy than in the first embodiment.

  Also in the second embodiment, as in the first embodiment, when video bit rate conversion is performed by the transcoder 3, the image format may be the same as the input and output, or when the bit rate falls below a predetermined threshold value. May be changed to a low resolution screen.

  Further, as in the first embodiment, by controlling the transcoder 3 with the set conversion bit rate based on the maximum value among the plurality of average bit rates Br, there is an effect that a more stable operation can be expected.

  As an application example of the present invention, the present invention can be applied to a video server apparatus that supplies a stream including video to a receiving apparatus via various networks.

  1 hard disk drive (HDD), 2 read control means, 3 transcoder, 4 stream generation means, 5 HTTP server, 6 network interface (I / F), 7 conversion control means, 8 sending packet monitor means, 9 stream buffer.

Claims (1)

A video stream transmission method for performing HTTP communication between a video server device and a receiving device connected via a network, establishing communication by TCP / IP, and transmitting data,
Reading the video stream from the video server device;
Convert the bit rate related to the read video stream to the set conversion bit rate,
For the video stream converted to the set conversion bit rate, a transmission packet including the position information of the stream to be transmitted to the network is generated,
A method for transmitting the packet for transmission to the receiving device;
The set conversion bit rate is an average transmission bit rate obtained based on the position information included in each of the plurality of sending packets and a time difference at which each sending packet is sent. Is less than the average transmission bit rate , it is controlled to be less than ,
Video stream transmission method.

Images