CN101563886B

CN101563886B - Redundant media packet streams

Info

Publication number: CN101563886B
Application number: CN200780026677.0A
Authority: CN
Inventors: 艾丹·威廉斯; 安德鲁·怀特
Original assignee: National ICT Australia Ltd
Current assignee: Aolian Holdings Pte. Ltd.
Priority date: 2006-05-17
Filing date: 2007-05-17
Publication date: 2012-12-05
Anticipated expiration: 2027-05-17
Also published as: CN101491002A; CN101491002B; CN101563886A

Abstract

This invention concerns the transmitting and receiving of digital media packets, such as audio and video channels and lighting instructions. In particular, the invention concerns the transmitting and receiving of redundant media packet streams. Samples are extracted (556) from a first (904) and second (906) media packet stream. The extracted samples are written to a buffer (910) based on the output time of each sample (556). Extracted samples having the same output time are written to the same location in the buffer. Both media packet streams are simply processed all the way to the buffer without any particular knowledge that one of the packet streams is actually redundant. This simplifies the management of the redundant packet streams, such as eliminating the need for a ''fail-over'' switch and the concept of an ''active stream''. The location is the storage space allocated to store one sample. The extracted sample written to the location may be written over another extracted sample from a different packet stream previously written to the location. These extracted samples written to the same location may be identical.

Description

Redundant media packet streams

Technical field

The present invention relates to transmission and reception to the grouping of the Digital Media such as Voice & Video passage and luminous instruction.These media channels are sent to receiver device as media packet from transmitter apparatus, to be used for broadcast.Particularly, the present invention relates to transmission and reception to redundant media packet streams.The computer software that the present invention relates to the method for transmitter apparatus, receiver device, data network, receiving media packet streams and carry out this method.

Background technology

For a long time, use the cable of application specific that the media channel such as the Voice & Video passage is transmitted.For example, use the double-core speaker cables that a left side and right voice-grade channel are sent to loud speaker from amplifier.

Recently, use the agreement such as clean culture or multicast to come the transfer medium signal through the computer based network.Clean culture is to send the method for dividing into groups through computer network to single goal ground.Hope to receive the media device that divides into groups to each, must retransmit unicast packet.

Typically, multicast refers to ip multicast, and ip multicast is an agreement of utilizing multicast address on the TCP/IP network, to send effectively to a plurality of receiver devices simultaneously.Computer network then operation receives on the equipment of multicast grouping to route the packet in each hope on the network.

Media network can allow to send and receive redundant media packet streams.Known, transmitter is handled main media packet streams continuously, and ignores the redundant copies of media packet streams simply.In main stream of packets, detect (that is, the transmission path in the network breaks off) in the incident of problem, " fault recovery " switch occurs.In case activate this switch, receiver device is ignored main media packet streams, and handles redundant copies.

Summary of the invention

On the one hand, the invention provides and be used to receive the receiver device from the media packet streams of data network, this receiver device comprises:

First data-interface is used to receive first media packet streams, and said first media packet streams comprises the sampling of media channel;

Second data-interface is used to receive second media packet streams, and said second media packet streams comprises the sampling of media channel;

Processor is used for extracting sampling from first media packet streams and second media packet streams, and the output time of definite sampling of being extracted;

Buffer is used for the sampling that temporary transient storage is extracted, to be used for output; And

Wherein, thus said processor operations based on the output time of correspondence with in the sampling write buffer that is extracted, so that the sampling with identical output time of being extracted is written into the same position in the buffer.

Like this, receiver device can be actually under the redundant situation and handle two media packet streams all the time simply in its write buffer not knowing one of stream of packets especially.This has just simplified the management to redundancy packets stream, has for example eliminated the demand for " fault recovery " switch.In addition, the notion that no longer needs " initiatively stream ".

Buffer can be designed as the output time sampling of being extracted of storing predetermined maximum number in order according to each sampling of being extracted.The position is the memory space that branch is used in a sampling of storage.First of writing position extracts sampling and can the second extraction sampling from different grouping stream that before write this position be rewritten.The first and second extraction samplings that write same position can be identical.

Buffer can be associated with an output channel, and each media packet streams can be directed to identical output channel.Each media packet streams can comprise two or more media channels.

Receiver device also can comprise the 3rd data-interface, is used to receive the 3rd media packet streams of the sampling that comprises media channel, and wherein, thereby said processor is further operated the output time of the sampling of from the 3rd media packet streams, extracting sampling and confirming to be extracted.Owing to need not be used to handle the cura specialis control of redundancy packets stream, therefore can increase the number of redundancy packets stream at an easy rate, and utilize same way as that all stream of packets are handled.

First and second media packet streams can receive from the different pieces of information interface of the identical transmitter apparatus that is connected with data network.

Processor also can be operated to detect ruined sampling and to prevent that these ruined samplings are written in the buffer.

All samplings that comprised in the media packet of first media packet streams can all not be comprised in any single medium grouping of second media packet streams.Even without in the same manner packetizing being carried out in the sampling of two Media Streams, processor also can be operated to use same way as to extract and write sampling.

The sampling of media channel can have related timestamp, and processor can be operated the output time with the sampling of confirming based on corresponding timestamp to be extracted.This timestamp can be that absolute time is stabbed, and can represent the sampling time.

The present invention can also comprise and be used to receive the method from the media packet streams of data network that the method comprising the steps of:

Receive first media packet streams, said first media packet streams comprises the sampling of media channel;

Receive second media packet streams, said second media packet streams comprises the sampling of media channel;

From said first media packet streams and second media packet streams, extract sampling;

The output time of definite sampling of being extracted;

Based on corresponding output time, with in the sampling write buffer that is extracted to be used for output, will have the same position in the sampling write buffer that is extracted of identical output time thus.

On the other hand, the invention provides computer software and carry out said method with the operation receiver device.

Aspect another, the invention provides and be used for sending the transmitter apparatus of media packet streams to receiver device through data network, this transmitter apparatus comprises:

One or more data-interfaces are used for sending first media packet streams and second media packet streams to receiver device, and said Media Stream comprises one or more media channels;

Processor carries out packetizing based on the request from receiver device to media channel, is used for the media packet streams that sends through data-interface with establishment; And

Controller makes said processor create two independently media packet streams, said two independently media packet streams comprise the identical media passage that will be sent to receiver device.

This request can be addressed to first media packet streams that comprises media channel first interface of receiver device, and second media packet streams that will comprise the identical media passage is addressed to second interface of receiver device.This request can be the single message that receives from receiver device.

Transmitter apparatus can comprise and is used to second data-interface that sends first data-interface of first media packet streams and be used to send second media packet streams.

Data network can comprise the first and second data sub-networks.Controller can be so that first media packet streams sends through the first data sub-network, and makes second media packet streams carry out through second sub-network.Like this, if a data network can't successfully transmit one or more groupings, then transmit identical media channel, and loss of data can not take place via another sub-network.In addition, the first data sub-network can have and the different configuration of the second data sub-network, for example different transport protocols.

The invention still further relates to the method and the software that are used for two media packet streams that comprise the identical media passage are sent to receiver device.

Aspect another, the invention provides the computer network that comprises above-mentioned receiver device and transmitter apparatus.

Description of drawings

Referring now to accompanying drawing example of the present invention is described, in the accompanying drawings:

Fig. 1 is the sketch map of the network that can together use with the present invention;

Fig. 2 is to use the sketch map of the single grouping of the media packet streams that sends through network of the present invention;

Fig. 3 (a) is in data network, how to realize redundant sketch map to 3 (d);

Fig. 4 is the sketch map that uses the network of redundant interface according to embodiment of the present invention;

Fig. 5 is another sketch map that another embodiment according to the present invention is used the network of redundant interface; And

Fig. 6 is the sketch map of buffer of the output channel of receiver device.

Specific embodiments

The networking component summary

With reference to Fig. 1, data network has been described wherein.Data network 100 comprises transmitter apparatus 110 and receiver device 112.In network 100, possibly there are a plurality of transmitter apparatus 110 and a plurality of receiver devices 112, but for the sake of clarity, one of them only are discussed here.In addition,

equipment

110 and 112 possibly be able to carry out sends and receives two functions, but here for for the purpose of clear more,

equipment

110 and 112 is described as only carrying out separately a function.

Transmitter apparatus 110 is connected to each other through network 114 with receiver device 112, so they can send and receive the Digital Media grouping as the part of media packet streams.Transmitter apparatus 110 comprises Audio Processing engine (APE) 120 and Audio Processing engine controller (APEC) 122.Receiver device 112 also comprises APE 124 and APEC 126.Media packet is sent in the media packet streams that can comprise one or more media channels.For simplicity, embodiment is described as identical form with all media channels in the media packet streams, yet the present invention can be applicable to a plurality of media channel forms in the media packet streams.

Media device 140 such as guitar is connected with APE 120, the media signal that APE 120 receiving media equipment 140 are produced.Can add the preprocessor (not shown), so that media channel is served as numeral or is transformed into another number format (for example, sampling rate or bit-depth conversion) from a kind of number format from analog-converted.Then, 120 pairs of digital media channels of APE carry out packetizing.Utilize network 114 stream of packets that produces to be sent to the APE 124 of receiver device 112.Then, 124 pairs of digital media signal of APE are separated packetizing, and (if suitably) is converted into analog signal, and the analog media signal is sent to media device 142, for example is used to the loud speaker that broadcasts.When media signal is the non-dummy source such as the MIDI source, needn't change.The speed and the skew of strict in time control packetizing and transmission, with guarantee through 142 pairs of media signals of media device broadcast be connected to same another media device (not shown) from APE 120 receiving media signals of network 114 to synchronous between the broadcast of media signal.In common unsettled PCT application PCT/AU2006/000538 (WO2006/110960), describe the packetizing operation of

APE

120 and 124 in detail.In addition, the also reference common unsettled PCT application of the priority that requires AU2006906015 and AU2006902741 of submission on the same day.

APEC 122/126 is with software or hard-wired assembly.In this network 100, APEC122/126 is on the identical physical equipment with APE 120/124, but can for example can be on another equipment or central computer that is connected with network 114 away from APE120/124 alternatively.APEC122/126 provides the general introduction of APE 120/124 and connected any

audio frequency apparatus

140 and 142 to the user.But transmitter apparatus 110 has a plurality of sendaisles of can being named, also can offering the receiver device 112 on the network 114; This is called as advertisement.Receiving equipment 112 has a plurality of receive paths.Can named sendaisle be distributed to receive path; This is called as reservation.APEC 122 and 126 will be configured APE 120/124, so that media signal is routed to receive path from sendaisle.

Receive and send APEC 126 and 122 through network 114 exchange configuration information and control messages.Via the 118 exchange configuration informations of the service discovery database such as DNS-SD.This database can utilize distributed way to realize, makes each

equipment

110 and 112 store and provide the configuration information that is associated with its APEC 122/126.Sometimes need add control messages and accomplish booking process, and make media signal be routed.These additional control messages are receiving APEC 126 and are sending between the APEC 122 and send.

Himself APE 120/124 of each APEC 122/126 configuration, and carry out alternately with other APEC 122/126 is to mate between the APE 120/124 that guarantees to be configured in communication.

In each

equipment

110 and 112 inside; Input channel such as from the passage of audio frequency apparatus 140 is called as " TX passage "; Because they will send through network; And the output channel such as the passage that is sent to audio frequency apparatus 142 is called as " RX passage ", because they will receive data from network.

At first, equipment must connect to create network 104.With reference to Fig. 4, four media devices (keyboard 144, guitar 140, be used for the left passage 146 and the right passage 148 that is used for CD of CD) are connected with four input channels of transmitter apparatus 110.In this example, keyboard uses identical sampling rate and sample format (such as 48kHz, 24 bits, pcm encoder) with guitar.The CD passage has different sampling rates and sample format (such as 44.1kHz, 16 bits, pcm encoder).

Media packet streams

Referring now to Fig. 2, it schematically shows single medium and divides into groups 550.A plurality of media packet that order is sent have constituted media packet streams.Each media packet 550 in the identical media stream comprises identical media channel.Each media packet 550 comprises frame 554.In every frame 554 of media packet 550, distribute sample space 556 must for each media channel.This sample space also is called as time slot.Each grouping 550 has time slot of every passage, has 554 1 sample spaces 556 of every hardwood thus.

Each frame 554 stabs with absolute time and is associated, and this absolute time is stabbed and also is recorded in the grouping.The notion of time is synchronous between all devices on the network 100 110 and 112, so their share common clock.The all devices 112 of the broadcast of the media channel that is comprised in the control frame 554 must synchronously broadcast the passage in each frame.This is based on, and timestamp in each frame 554 accomplishes.For example, with reference to common clock, each playing device can broadcast the sampling 556 in each frame 554 with the predetermined delay that surpasses the absolute time stamp.

Redundant media packet streams

Referring now to Fig. 4 redundancy is described.Redundancy is to realize through on the distinct interface of the network equipment, stream of packets being duplicated.Mainly redundancy is controlled at the APEC layer.APEC can be programmed to the given stream of packets on APE and will send or receive through the particular data interface.If send the specific medium stream of packets redundantly, APEC understands two or more copies of programming stream of packets so.Copy sends from first interface 130 of APE 120, and second copy of stream of packets sends from second interface 136 of APE 120.

Similarly, the APEC 126 that receives is programmed to its APE 124 through each data-

interface

132 and 134 and receives copy.Each stream of packets of duplicating is programmed to sampling is offered identical output channel.This means that stream of packets is written in the identical buffer.Owing to the sampling in two stream of packets is identical with determined output time, therefore identical sampling is handled, and override buffer, to be used for output channel with it.Usually, this state that will lead to errors and compete.Yet; Because all redundant media packet streams all comprise the identical sampling with synchronous output time; Therefore which sampled data of which media packet streams at first is written into buffer and not serious; Because all copies that repeats are override simply, thereby in fact a portion copy with each sampling sends to output channel.

Output time indication sampling will be write which position in the buffer.The output time of sampling be based on media packet streams in the timestamp that is associated of sampling and definite.This can comprise the delay of each media packet streams and skew are compensated.Note, need not add the physics timestamp for each sampling interpolation timestamp of media channel to each sampling.If in a grouping, send 20 samplings, then add timestamp to first sampling simply and also implied remaining interpolation timestamp from the cycle of single medium signal voice data.Therefore, remaining sampling also has timestamp through association.

Fig. 3 shows and on data network, realizes some redundant modes.Interface can be connected to independently network 114 (a) and 114 (b) or be connected to single network 114 via a plurality of paths.Interface 138 and 139 among Fig. 8 (c) is respectively second redundant interface of transmitter 110 and receiver 112.The interface that Fig. 8 (d) shows on the transmitter 110 can send by a plurality of separation interfaces on receiver 112.

APE

120 and 124 with a plurality of channel data interfaces specifies an interface as main interface separately.This main interface is interface #0.Any other (redundancy) channel data interface is from 1 open numbering.For example, the APE 124 that has a main interface 132 and a redundant interface 134 among Fig. 4 has interface #0 and #1.

For simplicity, when sending or receive,

APE

120 and 124 only is configured to and is equal to interface and sends or receive.Main interface 130 on APE 120 communicates with the main interface 132 on another APE 124.Interface 136#1 on APE 120 only with another APE 124 on interface 134#1 communicate.By that analogy.This just makes it possible to make marks from the outside to each APE interface.For example, hardware originally on one's body, main interface can black, first redundant interface is red, second redundant interface be a blueness, or the like.This just makes the user can easily distinguish different interfaces, and guarantees in their quilt wiring correctly of assembly process.

Redundancy in the unicast protocol

On network 114, media channel is carried out advertisement, can utilize unicast protocol to subscribe this media channel with the indication receiver.Receiver can be asked the TXT record of receiving media passage, and this TXT record comprises the information of media channel, comprises sampling rate (rate), bit-depth (bits) and coding (enc) (1=PCM)." txtvers " field is the version number of TXT record." nchan " field is the largest passages number of each dynamic beam.Each dynamic beam of interface box (stage-box) 110 has maximum four passages, and this is enough to send all inputs 144,140,146 and 148 through single grouping.The id field is any physical gap marker symbol that is used for identifying briefly its passage by APEC 120.The passage of " keyboard (keyboard) " by name just in time has ID 16.Except sending APEC 120, this ID is only helpful to the dynamic beam on the receiver APEC 124 configuration APEC 122.

The TXT record of advertisement comprises the field that is marked as " nred ".The value of " nred " has been indicated the number of redundant interface.If omit this field, then this value is regarded as zero.The null value indication does not have redundant interface, this means that the APE 120 of transmitter 110 only supports primary traffic.Value 1 indication has single redundant interface.Indicate a plurality of redundant interfaces (being numbered 1..n) greater than 1 value.

Except sending single dynamic beam request, the APEC 126 that receives can also send a request to redundant interface available on each transmitter, and each has different interface field.Alternatively, can the request to redundancy stream to redundant interface sent out be merged to and send out in the initial request message to stream of packets of nonredundancy interface.

As shown in Figure 4, except main interface 130 (#0), interface box 110 also has redundant interface 136 (#1).TXT record to the passage advertisement formats as follows now:

Record: keyboardstage-box._netaudio_chan._udp.local TXT

txtvers＝2

rate＝48000

bits＝24

enc＝1

nchan＝4

id＝16

nred＝1

Blender 112 is also supported the redundant channel.The address of its master data interface 132 is 169.254.28.12.The address of its second data-interface 134 is 169.254.132.15.

In this locality, blender 112 sends as follows ' creating the RX bundle ' message from its APEC 126 to its APE 124:

Field	Value	Remarks
			Destination address	169.254.28.12	The master data interface IP address of blender
Destination interface	26528
			Interface	0	The master data interface is #0
Number of active lanes	2
			Mapping for time slot 1	[8]	Array with an element: RX passage 1
Mapping for time slot 2	[11]	Array with an element: RX passage 4

Field	Value	Remarks
			Destination address	169.254.132.15	The second data-interface address of blender
Destination interface	28452
			Interface	1	Second data-interface is #1
Number of active lanes	2
			Mapping for time slot 1	[8]	Array with an element: RX passage 1
Mapping for time slot 2	[11]	Array with an element: RX passage 4

Each message disposes the stream of packets of a reception to interface 132 or 134.Second interface 134 (#1) can use the port numbers identical or different with main interface 132, and this depends on the design of APE 124.This example hypothesis is selected different ports.

Must on its APE 124, create two separated funiculuses as blender 112, it need create the dynamic beam of two separation on interface box 110.Following message is sent to APEC 122 from APEC 126:

Field	Value	Remarks
			Destination address	169.254.28.12	The master data interface IP address of blender
Destination interface	26528
			Interface	0	The master data interface is #0
Number of time slot	2
			Time slot 1	16	The I D of keyboardstage-box
Time slot
	2	17	The ID of guitarstage-box

Field	Value	Remarks
			Destination address	169.254.132.15	The second data-interface address of blender
Destination interface	28452
			Interface	1	Second data-interface is #1
Number of time slot	2
			Time slot 1	16	The I D of keyboardstage-box
Time slot
	2	17	The I D of gu itarstage-box

Finally, interface box 110 is created two bundles on local APE 120, to satisfy these requests.Following message is sent to APE 120 from APEC 122:

Field	Value	Remarks
			Destination address	169.254.28.12	The master data interface IP address of blender
Destination interface	26528
			Interface	0	(on the interface box) master data interface is #0
Number of active lanes	2
			The TX passage of time slot 1	16	TX passage 1
The TX passage of time slot 2	17	TX passage 1

Field	Value	Remarks
			Destination address	169.254.132.15	The second data-interface address of blender
Destination interface	26452
			Interface	1	(on the interface box) second data-interface is #1
Number of active lanes	2
			The TX passage of time slot 1	16	TX passage 1
The TX passage of time slot 2	17	TX passage 2

Multicast is redundant

In order to support redundant multicast, transmitter 110 is created some different packets streams, and with its as with separating of single bundle name of static Shu Jinhang advertisement.

For example, for two copies of advertisement bundle b1 and b2, promptly main beam 130 and second restraints 136, and interface box 110 will at first be stated the multicast address of each bundle.Main beam uses 239.254.46.46.Second bundle uses 239.254.98.147.For this example, suppose that two bundles all use identical interface (29061).

Create two service logs (SRV), record of each stream of packets.Owing to each stream of packets is carried out identical format, therefore only needs single TXT record.

Record: b1stage-box._netaudio_bund._udp.local SRV

0?1?2906146.46.254.239.mcast.local

Record: b1stage-box._netaudio_bund._udp.local SRV

1?1?29061?147.98.254.239.mcast.local

Record: b1stage-box._netaudio_bund._udp.local TXT

txtvers＝1

rate＝48000

bits＝24

enc＝1

nchan＝4

If only use single (master) interface, then " priority " field (being expressed as first number) among the SRV is set to 0.The indication of non-zero priority is applied to redundant interface with bundle, is interface 1 in this example.

These bundle advertisements are decoded and made receiver can on each interface, dispose suitable bundle.APEC 126 is configured to APE 124 on first main interface 132, to receive main beam, and on second interface 134, receives second bundle.

Buffer management

No matter use clean culture or multicast protocol, the mode of the stream of packets that receiver device 112 administrative institutes duplicate is identical.

With reference to the data network among Fig. 5 902, it shows transmitter apparatus 110 and sends two identical stream of packets.A stream of packets 904 is sent autonomous data-interface 130, the second stream of packets 906 and is sent from first redundant interface 136.Two stream of packets all comprise the sampling that media channel produced that is connected with keyboard 144.

The first main stream of packets 904 is received at master data interface 132 places by receiver device (being amplifier 908).Second redundancy packets stream 906 is received at the first redundant data interface, 134 places by amplifier 908.APE 124 handles stream of packets 904 and 906 each according to the instruction that receives from APEC 126.

Here, APE 124 will handle the stream of packets 904 and 906 that receives, and to broadcast through loud speaker 142, loud speaker 142 is connected to the output channel 2 of amplifier 908.Alternatively, the equipment that is connected to output channel 2 can not be playing device, can be for example blender or amplifier.

APE 124 extracts sampling from the data flow 904 and 906 that receives, and stabs to confirm the output time of each sampling based on time corresponding.Then, sampling is being sent to output channel 2 with before broadcasting through loud speaker 142, APE 124 is based on the determined output time write buffer 910 of will sampling.

The method of the write buffer of will sampling is described referring now to Fig. 6.Processing packet flows 904 and 906, and APE independently 124 does not need to recognize especially that they are identical in fact.

APE 124 extracts the sampling of keyboard media channel from stream of packets, and it is write in 912 buffers 910, so that sampling is in the output order in buffer.Conceptive buffer is divided into t at interval ₀, t ₂, t ₃... t _n, it is as the length of each sampling, and wherein each all corresponding to an output time.Output time based on separately writes a sampling in each interval.In case some samplings are written into last the interval t in the buffer 910 _n, continue sampling is write the first interval t ₀, with covering write wherein but be transferred into output channel 2 and by loud speaker 142 broadcast before sampling.

Because each sampling that receives all has related timestamp, this timestamp has the absolute time benchmark, therefore can easily confirm and the corresponding output time of absolute time benchmark.In this example, must with identical sample-synchronous be sent to output channel, so the output time of each sampling also must be identical.Therefore,, then also can it be write correct interval, because can select correct interval based on the output time of this sampling if some samplings do not receive in order.If some samplings belong to the time interval of having been broadcasted, then will lose this sampling simply.

First stream of packets 904 and second stream of packets 906 all utilize this mode to handle.Because sampling and the output time of confirming thereof are identical, therefore identical sampling is write same interval by twice.Earlier will be from the sampling of first stream of packets 904 or earlier will be unimportant from the sampling write buffer 910 of second stream of packets 906, the final result that temporarily is stored in the sampling in the buffer is all with identical.

Like this, if lost during the transmission that is grouped in transmitter apparatus 110 from first stream of packets 904, so will sampling write buffer 910 from the same packets in second stream of packets 906 in.In this case, only with these the sampling write buffers in once.Certainly, from the packet loss of second stream of packets 906 time, also opposite situation possibly take place, and in only will the sampling write buffer 910 from the same packets in first stream of packets 906.

Because APEC 126 knows stream of packets 904 and 906 (more specifically being the particular time-slot in each stream of packets 904 and 906) and comprises identical data; And be sent to the message of APE 124 from APEC 126 above-mentioned; APEC 126 is programmed to APE 124 unconditionally both copies is advanced same buffer, and APE 124 need not judge that which stream of packets is " current " or " effectively ".Handle on the same intervals (position) of APE 124 in buffer and copy the two, and need not consider earlier in stream of packets 904 and 906 which handled and with in its write buffer, two streams are all with the common mode simple process.This has reduced the complexity that the system that can send and receive redundancy packets stream is managed.For example, no longer need judge whether activate the fault recovery switch.In case set up the path, only if all fail in two paths, otherwise APE 124 or APEC 126 do not need other decision logic.

Be destroyed if during Network Transmission, sample, then identification in a single day, those samplings should be lost.The media packet streams that duplicates will guarantee that output buffer receives at least one copy of sampling, and need not between media packet streams, to switch.Before sampled data being write shared buffer, carry out any transport level error-tested or mistake and recover.

The actual samples content of each grouping in the media packet streams 904 and 906 does not need identical.For example, can by different way media channel be divided into groups to turn to stream of packets 904 and 906, so that do not comprise identical sampling from the grouping of each stream.The method of sampling in the write buffer is identical all the time, because regardless of the position of the disassociation frame in dividing into groups, each identical sampling all will be shared output time.In the time of in the write buffer of will sampling, which (position) at interval APE 124 will should be written into to confirm this sampling with reference to the output time of each sampling.

The present invention also allows other redundant stream of each self-contained identical media channel.Can each media packet streams be sent to different data-interfaces.Same, APE 124 will handle all stream of packets independently, and the sampling of being extracted is write in the identical buffer, to be used for output channel.

Certainly, sampling 556 can be the mixing of one or more source of media, for example keyboard 144 and guitar 140.Alternatively, blender can be provided, it be mixed before the write buffer will sampling being used for.

It will be understood by those skilled in the art that under the prerequisite that does not deviate from broadly described the spirit or scope of the present invention, can carry out multiple variant and/or modification the present invention shown in the specific embodiments.

■ when using static bundle, can with arbitrarily easily mode give bundle with channel allocation.Also can realize redundancy through on special interface, creating specific bundle.

The ■ dynamic beam is generally clean culture.Static bundle is generally multicast.If necessary, also can dynamic beam be configured to multicast, perhaps the static state bundle be configured to clean culture.

The I/O of ■ physical channel also can realize with form of software, for example with the form of software of blender, and needs not to be the actual physical connector.

Although the above-mentioned example of ■ has been supposed the channel data of sampling, yet identical protocol will work to the cycle data stream of any fixed size.Utilize the variant of transmission (dividing into groups and bundle) mechanism, also can support the data of aperiodic or variable-size.

It is in order to realize facility that ■ gives time slot allocation isomorphism passage.Can utilize start byte and length in frame, to identify to have the non-isomorphism passage media data of common sampling rate.

■ MIDI is the example of non-cycle data.The same-interface of route audio frequency to Voice & Video to video can route MIDI to MIDI abstractively to make the user.The same with the cycle grouping, MIDI divides into groups to have timestamp, but receiver should not expect that the MIDI data cell will periodically arrive.The aperiodicity data possibly need periodic " keepalive " message to distinguish the nand function stream of mourning in silence.As long as there is the mark of explanation " this divide into groups in not these data ", then the aperiodicity data can the same auto-polymerization easily with periodic data.

Therefore, existing embodiment all should be considered to exemplary in all respects and and nonrestrictive.

Claims

1. be used to receive the receiver device from the media packet streams of data network, said receiver device comprises:

Second data-interface is used to receive second media packet streams, and said second media packet streams comprises the said sampling of said media channel;

Processor is used for extracting sampling from said first media packet streams and said second media packet streams, and confirms the output time of the sampling of this extraction;

Buffer is used for temporarily storing the sampling of said extraction, to be used for output; And

Wherein, the sampling with said extraction writes in the said buffer thereby said processor operations is based on said output time, and the sampling that has the extraction of identical output time thus is written into the same position in this buffer.

2. receiver device according to claim 1, wherein, the sampling of the extraction that said buffer can temporary transient storing predetermined maximum number.

3. receiver device according to claim 1 and 2, wherein, said buffer can temporarily be stored the sampling of this extraction in order according to the corresponding output time of the sampling of said extraction.

4. receiver device according to claim 1 and 2, wherein, said output time is the broadcast time that media device broadcasts said sampling.

5. receiver device according to claim 1, wherein, said position is the memory space that branch is used in a sampling of storage.

6. receiver device according to claim 1 wherein, is written into the sampling of extracting from second of different packets stream that the sampling of extracting first of said position writes this position before rewriting.

7. receiver device according to claim 6, wherein, the sampling that the sampling of said first extraction and said second is extracted is identical.

8. receiver device according to claim 1, wherein, said buffer is associated with an output channel, and each media packet streams is directed to identical output channel.

9. receiver device according to claim 1; Wherein, said receiver device also comprises the 3rd data-interface, is used to receive the 3rd media packet streams of the said sampling that comprises said media channel; Wherein, Said processor is also operated from the 3rd media packet streams, extracting sampling, and confirms that output time and the sampling that will extract based on this corresponding output time of the sampling of this extraction write in the said buffer.

10. receiver device according to claim 1, wherein, said processor is also operated detecting ruined sampling, and prevents that these ruined samplings are written in the said buffer.

11. receiver device according to claim 1, wherein, all samplings that comprised in the media packet of said first media packet streams are not all to be comprised in any single medium grouping of said second media packet streams.

12. receiver device according to claim 1, wherein, said first media packet streams and said second media packet streams are that the different pieces of information interface from the transmitter apparatus that is connected with said data network receives.

13. receiver device according to claim 1, wherein, the said sampling of said media channel has related timestamp separately, and said processor operations is to confirm the output time of the sampling of this extraction based on the timestamp of said association.

14. be used to receive the method from the media packet streams of data network, said method comprises step:

Receive second media packet streams, said second media packet streams comprises the said sampling of said media channel;

From said first media packet streams and said second media packet streams, extract sampling;

Confirm the output time of the sampling of said extraction;

Based on the corresponding output time of the sampling of said extraction, the sampling of each said extraction is write in the said buffer to be used for output, the sampling that has the said extraction of identical output time thus is written into the same position in this buffer.

15. method according to claim 14, wherein, the step that said sampling with each said extraction writes in the said buffer comprises: write the sampling of said extraction in order according to the corresponding output time of the sampling of each said extraction.

16. method according to claim 14, wherein, said output time is the broadcast time that media device broadcasts said sampling.

17. method according to claim 14, wherein, said position is the memory space that branch is used in a sampling of storage.

18. method according to claim 14; Wherein, The step that said sampling with each said extraction writes in the said buffer comprises: the sampling of first extraction is write a position, use the sampling of extracting from second of different grouping stream to rewrite the sampling of this first extraction of said position then.

19. method according to claim 18, wherein, the sampling that the sampling of said first extraction and said second is extracted is identical.

20. method according to claim 14, wherein, said buffer is associated with an output channel, and each media packet streams is directed to identical output channel.

21. method according to claim 14, wherein, said method is further comprising the steps of: receive the 3rd media packet streams of the said sampling that comprises said media channel, and from said the 3rd media packet streams, extract sampling.

22. method according to claim 14, wherein, said method is further comprising the steps of: detect ruined sampling, and make these ruined samplings not be written in the said buffer.

23. method according to claim 14, wherein, all samplings that comprised in the media packet of said first media packet streams are not all to be comprised in any single medium grouping of said second media packet streams.

24. method according to claim 14, wherein, the sampling in the said media channel has related timestamp separately, and the step of the output time of the sampling of said definite said extraction is based on the timestamp of said association.

25. method according to claim 14; Wherein, Said first media packet streams is from first data-interface of the transmitter apparatus that is connected with said data network, and said second media packet streams is from said first data-interface or another data-interface of the transmitter apparatus that is connected with said data network.

26. method according to claim 14, wherein, said method comprises that also sending a message asks to receive said first media packet streams and said second media packet streams.

27. data network comprises:

Be used for through the transmitter apparatus of data network to receiver device transmission media packet streams, said transmitter apparatus comprises:

One or more data-interfaces are used for sending first media packet streams and second media packet streams to said receiver device, and said media packet streams comprises one or more media channels;

Processor carries out packetizing based on the request from said receiver device to said media channel, is used for the media packet streams that sends through said data-interface with establishment; And

Controller makes said processor create two independently media packet streams, said two independently media packet streams comprise the identical media passage that will be sent to said receiver device;

According to each described receiver device in the claim 1 to 13; And

Communicator is used to make media packet streams to be sent to said receiver device from said transmitter apparatus.