CN105071897A - Multipath redundant transmission method for network real-time audio conversation media data - Google Patents

Abstract

The invention provides a multipath redundant transmission method for network real-time audio conversation media data, and the method comprises the steps: enabling audio media data obtained by an audio media data transmitting end to be packaged according to an audio coding and decoding format and a network transmission protocol, which are negotiated by two sides of a real-time audio conversation; transmitting the data to a plurality of transmission paths, which are not all intersected with each other, at the same time in a mode of redundancy; carrying out the multipath redundant transmission of the real-time audio media data; carrying out the redundancy elimination and regrouping operation of the audio media data groups received from different transmission paths through an audio media data receiving end; and restoring the raw audio data according to the audio coding and decoding format and the network transmission protocol, which are negotiated by two sides of the real-time audio conversation. Therefore, the method can effectively reduce the overall impact on end-to-end transmission of media data from package loss, time delay and shaking caused by the condition change of a single path, thereby improving the reliability of the transmission of the real-time audio media data, and improving the quality of experience (QoE) of network audio conversation services.

Description

Audio session media data multi-path redundancy transmission method during a kind of network implementation

Technical field:

The present invention relates to network communication technology field, audio session media data multi-path redundancy transmission method when particularly relating to a kind of network implementation.

Background technology:

Internet adopts the transmission mechanism of " doing one's best ", service quality (QoS) can not be provided to ensure for end-to-end media delivery.Real-time audio session is the network traffic of a quasi-representative, and its media delivery takies the transmission bandwidth of a few kbps to hundreds of kbps usually, is a class narrow band service, although occupied bandwidth is little, higher to transmission requirement of real-time.Current IP communication terminal audio frequency media transmission usually adopts traditional RTP and udp protocol to carry out transmission and controls, the end-to-end default route path (single-pathway) that employing does not have service quality (QoS) to ensure is transmitted, in end-to-end path link, the uncertain congested data packetloss that causes and delay variation usually cause audio media data to recombinate and to decode, and have a strong impact on the quality of real-time audio session.Internet has become the important bearer network of network service, and improve real-time audio conversation media transmission means, promoting business experience quality is major issue urgently to be resolved hurrily.

Summary of the invention:

For the defect of prior art, audio session media data multi-path redundancy transmission method when the invention provides a kind of network implementation, the method takies the lower feature of transmission bandwidth in conjunction with real-time audio media, not exclusively transmission path is intersected end to end by building, and adopt multi-path redundancy to transmit controlling mechanism and agreement, realize the transmission of real-time audio media data multipath redundant fashion.This multi-path redundancy transmission method can effectively reduce single-pathway condition change cause packet loss, delay time and jitter is to the entire effect of the end-to-end transmission of media data, and then promote the reliability of real-time audio media data transmission, promote the user experience quality of network audio session service.

Audio session media data multi-path redundancy transmission method when the invention provides a kind of network implementation, comprising:

The audio coding decoding form that the audio media data of catching is consulted according to real-time audio session both sides by the transmitting terminal of audio media data and network transmission protocol packing, adopt redundant fashion to be sent on many transmission paths not exclusively intersected simultaneously, carry out the transmission of real-time audio media data multi-path redundancy;

The receiving terminal of audio media data carries out redundant eliminating and reorganization operation to the grouping of the audio media data from different transmission path received, and the audio coding decoding form consulted according to real-time audio session both sides and network transmission protocol reduction original audio data.

Alternatively, described many transmission paths not exclusively intersected comprise: a default path based on default route and one or more redundant transmission paths not exclusively intersected.

Alternatively, arrange redundancy coefficient according to described real-time audio session, the number of described many transmission paths not exclusively intersected is greater than the redundancy coefficient of described real-time audio session setting.

Alternatively, described redundant eliminating and reorganization operation, comprising:

Arrange and initialization subflow buffering area, reassembly buffer district and jitter elimination window;

The receiving terminal of described audio media data carries out redundant eliminating operation to the described audio media data grouping received;

The receiving terminal of described audio media data carries out reorganization operation to the described audio media data grouping after redundant eliminating operation.

Alternatively, described setting initialization subflow buffering area, reassembly buffer district and jitter elimination window, comprising:

Arrange the subflow buffering area with not exclusively crossing transmission path number same number, described subflow buffering area is for receiving the audio media data grouping of different transmission path, and each memory location of each described subflow buffering area of initialization is empty;

Reassembly buffer district is set, the sequence number of the audio media data grouping after described reassembly buffer district is used for storage redundancy rejecting, the sequence number storing the subflow buffering area of described audio media data grouping and described audio media data are grouped in the memory location in the subflow buffering area of storage, and described in initialization, reassembly buffer district size is N _jthe sequence number of the audio media data grouping after the redundant eliminating for storing in described reassembly buffer district operates all is initialized as-1, and the memory location that the sequence number of the subflow buffering area of described storage described audio media data grouping and described audio media data are grouped in the subflow buffering area of storage is all initialized as sky;

Arrange jitter elimination window, described jitter elimination window is for realizing the jitter elimination of described audio media data grouping, and the size of jitter elimination window described in initialization is W, W ∈ [W _min, W _max].

Alternatively, described redundant eliminating operation, comprising:

The mode of S1, employing poll inquires about each described subflow buffering area, obtains the up-to-date described audio media data grouping i received, extracts the sequence number of described audio media data grouping i, be designated as FSN _i;

S2, the sequence number FSN that described audio media data is divided into groups _iwith described reassembly buffer district size N _jcarry out modular arithmetic, obtain numerical value m, i.e. m=FSN _imodN _j, inquire about the sequence number of the audio media data grouping that m position in described reassembly buffer district stores, be designated as J_FSN _m;

If S3 is J_FSN _m=-1, then by FSN _ivalue be assigned to J_FSN _m, the sequence number of the subflow buffering area at described audio media data grouping i place and memory location are stored in m position, reassembly buffer district, perform step S1;

If S4 is J_FSN _m≠-1, FSN _i> J_FSN _m, then by FSN _ivalue be assigned to J_FSN _m, the sequence number of the subflow buffering area at described audio media data grouping i place and memory location are stored in m position, reassembly buffer district, perform step S1;

If S5 is J_FSN _m≠-1, FSN _i≤ J_FSN _m, then step S1 is performed.

Alternatively, the mode of described employing poll inquires about each described subflow buffering area, comprising:

The reception of the audio media data grouping in each described subflow buffering area of receiving terminal periodic statistical of described audio media data and redundant eliminating situation, obtain the polling priority sequence of subflow buffering area corresponding to different transmission path;

The receiving terminal of described audio media data, according to each described subflow buffering area of described polling priority sequence queries, carries out redundant eliminating operation.

Alternatively, described reorganization operation, comprising:

Dynamic adjustments jitter elimination window size;

According to described jitter elimination window size, carry out described audio media data grouping callback operation.

Alternatively, described dynamic adjustments jitter elimination window size, comprising:

Record audio media data packet i arrives time R (i) in described reassembly buffer district, according to timestamp or other flag bits for record audio media data packet transmitting time in described audio media data grouping i, obtain the transmitting time S (i) of described audio media data grouping i, calculate delay variation J (i)=R (the i)-S (i) of described audio media data grouping i;

According to the audio media data grouping arriving described reassembly buffer district, obtain the expection delay variation of the audio media data grouping i newly arriving described reassembly buffer district wherein, N is fixed value, and i>=N, P (k) they are weight coefficient, and $\underset{k=i-N}{\overset{i-1}{\Σ}}P\left(k\right)=1;$

Calculate the class standard error of described audio media data grouping i delay variation $J_{ARMSE}\left(i\right)=\sqrt{\frac{1}{N}-\underset{k=i-N+1}{\overset{i}{\Σ}}{\[J\left(k\right)-\stackrel{\‾}{J\left(k\right)}\]}^2};$

Threshold value g is set ₁, g ₂for judging jitter elimination window size change scope, and g ₁< g ₂;

When increase jitter elimination window size, if J _aRMSE(i) ∈ [g ₁, g ₂], jitter elimination window increases if J _aRMSE(i) > g ₂, jitter elimination window increases wherein, k ∈ [0.5,1];

When reduction jitter elimination window size, if J _aRMSE(i) ∈ [g ₁, g ₂], jitter elimination window reduces if J _aRMSE(i) > g ₂, jitter elimination window reduces wherein, k ∈ [0.5,1];

If the size W of the jitter elimination window after regulating is less than jitter elimination window minimum W _min, then the size of jitter elimination window is set to jitter elimination window minimum W _minif the size W of the jitter elimination window after regulating is greater than jitter elimination window maximum W _max, then the size of jitter elimination window is set to jitter elimination window maximum W _max;

Work as J _aRMSE(i) < g ₁, the size of jitter elimination window is without the need to changing;

If revise jitter elimination window size, then the expection delay variation of the audio media data grouping i changed for arbitrary integer in interval [J (i)-4, J (i)+4], otherwise, do not change.

Alternatively, described according to described jitter elimination window size, carry out described audio media data grouping callback operation, comprising:

The audio media data grouping that audio media data grouping serial number is FSN is found in jitter elimination window, the audio coding decoding form consulted according to real-time audio session both sides and the network transmission protocol carry out readjustment decoding, search the audio media data grouping that audio media data grouping serial number is FSN+1 in jitter elimination window;

If find, it is then the audio media data grouping readjustment decoding of FSN+1 by audio media data grouping serial number, continue to search the audio media data grouping that audio media data grouping serial number is FSN+2, otherwise, judge whether jitter elimination window has residue, if without residue, be then that the audio media data grouping of FSN+1 performs and once adjusts back decode operation by audio media data grouping serial number, continue to search the audio media data grouping that audio media data grouping serial number is FSN+2, otherwise, continue to search after waiting for preset time t;

If continuous three times are searched successfully, then jitter elimination window is reduced the size of a Frame, judge now whether jitter elimination window is less than jitter elimination window minimum W _min, if so, then the size of jitter elimination window is set to jitter elimination window minimum W _min, continue to perform search operation;

If continuous three times are searched unsuccessful, then jitter elimination window is increased the size of a Frame, judge now whether jitter elimination window is greater than jitter elimination window maximum W _max, if so, then the size of jitter elimination window is set to jitter elimination window maximum W _max, continue to perform search operation.

As shown from the above technical solution, audio session media data multi-path redundancy transmission method during network implementation of the present invention, comprise: the audio coding decoding form that the audio media data of catching is consulted according to real-time audio session both sides by the transmitting terminal of audio media data and network transmission protocol packing, adopt redundant fashion to be sent on many transmission paths not exclusively intersected simultaneously, carry out the transmission of real-time audio media data multi-path redundancy, the receiving terminal of audio media data carries out redundant eliminating and reorganization operation to the grouping of the audio media data from different transmission path received, and according to real-time audio session both sides consult audio coding decoding form and the network transmission protocol reduction original audio data.Thus, by effectively can improve the Discarded Packets compensation probability in transfer of data for real-time audio media data employing multipath redundant transmission, improve packet loss, delay time and jitter index that transmission is relevant, and then promote the reliability of transfer of data, improve business experience quality (QoE).

Accompanying drawing illustrates:

Audio session media data multi-path redundancy transmission method schematic flow sheet during the network implementation that Fig. 1 provides for first embodiment of the invention;

The multi-path redundancy transmission system structure chart of the multipath relay transmission service system of the SIPProxy/IMSCSCF participation session negotiation that Fig. 2 provides for second embodiment of the invention;

Audio session media data multi-path redundancy transmission method schematic flow sheet during the network implementation that Fig. 3 provides for second embodiment of the invention;

The transmission schematic diagram of the audio media data grouping of the multipath relay transmission service system of the SIPProxy/IMSCSCF participation session negotiation that Fig. 4 provides for second embodiment of the invention;

The message flow chart of the multi-path redundancy transmitting procedure of the multipath relay transmission service system of the SIPProxy/IMSCSCF participation session negotiation that Fig. 5 provides for second embodiment of the invention;

During the network implementation that Fig. 6 provides for second embodiment of the invention, the audio media data grouping rx-side buffering district of audio session media data multi-path redundancy transmission method arranges block diagram;

During the network implementation that Fig. 7 provides for second embodiment of the invention, design drawing is implemented in the buffering area of the audio media data grouping receiving terminal of audio session media data multi-path redundancy transmission method;

The encapsulation format schematic diagram of the audio media data grouping of the multi-path redundancy transmission of the multipath relay transmission service system of the SIPProxy/IMSCSCF participation session negotiation that Fig. 8 provides for second embodiment of the invention;

The redundant eliminating flow chart of the audio media data grouping of audio session media data multi-path redundancy transmission method during the network implementation that Fig. 9 provides for second embodiment of the invention.

Embodiment:

Below in conjunction with drawings and Examples, the specific embodiment of the present invention is described in further detail.Following examples for illustration of the present invention, but are not used for limiting the scope of the invention.

Audio session media data multi-path redundancy transmission method schematic flow sheet when Fig. 1 shows the network implementation that first embodiment of the invention provides, as shown in Figure 1, the method for the present embodiment is as described below.

101, the transmitting terminal of audio media data audio coding decoding form that the audio media data of catching is consulted according to real-time audio session both sides and network transmission protocol packing, adopt redundant fashion to be sent on many transmission paths not exclusively intersected simultaneously, carry out the transmission of real-time audio media data multi-path redundancy.

In this step, in audio session process of establishing, calling party and the callee of audio session obtain many not exclusively crossing transmission paths, it should be noted that described many transmission paths not exclusively intersected comprise: a default path based on default route and one or more redundant transmission paths not exclusively intersected.

Particularly, carry out in process at audio session, arrange redundancy coefficient according to described real-time audio session, the number of described many transmission paths not exclusively intersected is greater than the redundancy coefficient of described real-time audio session setting.

102, audio media data receiving terminal to receive the audio media data from different transmission path grouping carry out redundant eliminating and reorganization operation, and according to real-time audio session both sides consult audio coding decoding form and the network transmission protocol reduction original audio data.

In this step, described redundant eliminating and reorganization operation, comprising:

Arrange and initialization subflow buffering area, reassembly buffer district and jitter elimination window;

The receiving terminal of described audio media data carries out redundant eliminating to the described audio media data grouping received;

The receiving terminal of described audio media data carries out reorganization operation to the described audio media data grouping after redundant eliminating operation.

Further, described setting initialization subflow buffering area, reassembly buffer district and jitter elimination window, comprising:

Arrange the subflow buffering area with not exclusively crossing transmission path number same number, described subflow buffering area is for receiving the audio media data grouping of different transmission path, and each memory location of each described subflow buffering area of initialization is empty;

Reassembly buffer district is set, the sequence number FSN of the audio media data grouping after described reassembly buffer district is used for storage redundancy rejecting, the sequence number storing the subflow buffering area of described audio media data grouping and described audio media data are grouped in the memory location in the subflow buffering area of storage, and described in initialization, reassembly buffer district size is N _jthe sequence number FSN of the audio media data grouping after the redundant eliminating for storing in described reassembly buffer district operates all is initialized as-1, and the memory location that the subflow buffering area sequence number of described storage described audio media data grouping and described audio media data are grouped in the subflow buffering area of storage is all initialized as sky;

Arrange jitter elimination window, described jitter elimination window is for realizing the jitter elimination of described audio media data grouping, and the size of jitter elimination window described in initialization is W, W ∈ [W _min, W _max].

Further, the redundant eliminating operation of described audio media data grouping, comprising:

The mode of S1, employing poll inquires about each described subflow buffering area, obtains the up-to-date described audio media data grouping i received, extracts the sequence number of described audio media data grouping i, be designated as FSN _i;

S2, the sequence number FSN that described audio media data is divided into groups _iwith described reassembly buffer district size N _jcarry out modular arithmetic, obtain numerical value m, i.e. m=FSN _imodN _j, inquire about the sequence number of the audio media data grouping that m position, described reassembly buffer district stores, be designated as J_FSN _m;

If S3 is J_FSN _m=-1, then by FSN _ivalue be assigned to J_FSN _m, the sequence number of the subflow buffering area at described audio media data grouping i place and memory location are stored in m position, reassembly buffer district, perform step S1;

If S4 is J_FSN _m≠-1, FSN _i> J_FSN _m, then by FSN _ivalue be assigned to J_FSN _m, the sequence number of the subflow buffering area at described audio media data grouping i place and memory location are stored in m position, reassembly buffer district, perform step S1;

If S5 is J_FSN _m≠-1, FSN _i≤ J_FSN _m, then step S1 is performed.

It should be noted that the mode of described employing poll inquires about each described subflow buffering area, comprising:

The reception of the audio media data grouping in each described subflow buffering area of receiving terminal periodic statistical of described audio media data and the situation of rejecting, obtain the polling priority sequence of subflow buffering area corresponding to different transmission path;

The receiving terminal of described audio media data, according to each described subflow buffering area of described polling priority sequence queries, carries out redundant eliminating operation.

Further, described reorganization operation, comprising:

Dynamic adjustments jitter elimination window size;

According to described jitter elimination window size, carry out described audio media data grouping callback operation.

Particularly, carry out in process in the session of multipath real-time audio, the receiving terminal of voice data packet receives the audio media data grouping of described many transmission path not exclusively intersected, and the mark can distinguishing the grouping of described audio media data according to total sequence number or timestamp etc. in described audio media data grouping carries out the redundant eliminating operation of audio media data grouping; By searching the audio media data grouping in dynamic buffering district or dynamic buffering window, obtain audio media data grouping to be decoded, the audio coding decoding form consulted according to real-time audio session both sides and the network transmission protocol are carried out readjustment to the grouping of described audio media data and are decoded, and reduce original voice data.

The SIPProxy/IMSCSCF that Fig. 2 shows second embodiment of the invention to be provided participates in the multi-path redundancy transmission system structure chart of the multipath relay transmission service system of session negotiation, as shown in Figure 2, calling party 210 and callee 220 lay respectively at the Customer Premises Network at two ends, and SIPProxy/IMSCSCF230, media relays controller 240, media relay servers 250 are all deployed in network.

Described SIPProxy/IMSCSCF230 is the SIPProxy/IMSCSCF with multi-path transmission session negotiation ability.The signaling negotiation for multipath session transmissions is realized by expansion SIP and SDP message, in SIPProxy/IMSCSCF230, add a module for the treatment of multipath session request, effectively can realize the multipath session request of SIPProxy/IMSCSCF230 process from two ends Customer Premises Network.

Described media relays controller 240 is for the network topology of management of media Relay Server 250 and behavior, and be responsible for the distribution of relay transmission path, described media relay servers 250 participates in the structure of relay transmission path, is responsible for reception and the forwarding of packet.Described media relays controller 240 forms multipath relay transmission service system with described media relay servers 250, for multipath session provides relay transmission service, for example, two media relay servers are comprised, the media relay servers 250-1 namely shown in Fig. 2 and media relay servers 250-2 in the present embodiment.

Particularly, the SIPProxy/IMSCSCF230 having multi-path transmission session negotiation ability receives the multipath session establishment request of the Customer Premises Network from two ends, to media relays controller 240 request dispatching relay transmission path.Described media relays controller 240 and described media relay servers 250 have consulted distribution and the foundation of relay route, return the assignment information of relay transmission path to the SIPProxy/IMSCSCF230 with multi-path transmission session negotiation ability.

The SIPProxy/IMSCSCF of above-mentioned Fig. 2 participates in the multipath relay transmission service system of session negotiation, audio session media data multi-path redundancy transmission method schematic flow sheet during the network implementation that the second embodiment of the invention shown in composition graphs 3 provides, sets forth the method for the present embodiment further.

301, real-time audio session is set up between calling party and callee, the audio coding decoding form that the audio media data of catching is consulted according to real-time audio session both sides by the transmitting terminal of audio media data and network transmission protocol packing, adopt redundant fashion to be sent on three transmission paths not exclusively intersected simultaneously, carry out the transmission of real-time audio media data multi-path redundancy.

In this step, as shown in Figure 2, according to audio session, redundancy coefficient is set, the redundancy coefficient of the audio media data transmission in the present embodiment is 2, three transmission paths 260 are obtained, comprising a default path 260-D based on default route, via the relay transmission path 260-R1 of described media relay servers 250-1 and the relay transmission path 260-R2 via described media relay servers 250-2 between calling party 210 and callee 220.

The SIPProxy/IMSCSCF that Fig. 4 shows second embodiment of the invention to be provided participates in the transmission schematic diagram of the audio media data grouping of the multipath relay transmission service system of session negotiation, as shown in Figure 4, every transmission paths transmits identical audio media data grouping.

The SIPProxy/IMSCSCF that Fig. 5 shows second embodiment of the invention to be provided participates in the message flow chart of the multi-path redundancy transmitting procedure of the multipath relay transmission service system of session negotiation, and the signaling procedure of concrete multipath redundant transmission is as described below.

501, calling party 210 is to having the SIPProxy/IMSCSCF230 request of multi-path transmission session negotiation function for set up the session of multipath real-time audio between calling party 210 and callee 220, and the redundancy coefficient of audio session is 2;

The SIPProxy/IMSCSCF230 502, with Multipath Transmission session negotiation ability checks whether calling party 210 and callee 220 specifically have multipath session ability, if two sides do not have multipath session ability or a side not to have multipath session ability, the SIPProxy/IMSCSCF230 with Multipath Transmission session negotiation ability just refuses the request of multipath audio session.If calling party 210 and callee 220 have multipath session ability, the SIPProxy/IMSCSCF230 with Multipath Transmission session negotiation ability just asks as multipath real-time audio session between calling party 210 and callee 220 distributes many relay transmission path to media relays controller 240;

503, because the redundancy coefficient of session is 2, media relays controller 240 is that between calling party 210 and callee 220, the session of multipath real-time audio is assigned with two relay transmission path, article one, via media relay servers 250-1, one via media relay servers 250-2.Media relays controller 240 sends relay route respectively to media relay servers 250-1 and media relay servers 250-2 and adds request 503-1 and 503-2;

504, media relay servers 250-1 and media relay servers 250-2 completes the interpolation of relay route, returns relay route add success response 504-1 and 504-2 respectively to media relays controller 240;

505, media relays controller 240 returns relay route to the SIPProxy/IMSCSCF230 with multi-path transmission session negotiation ability and is allocated successfully response, comprises the information of two relay route in response message;

The SIPProxy/IMSCSCF230 506, with multi-path transmission session negotiation ability notifies the distribution condition of callee 220 transmission path;

507, callee 220 is multipath audio session distribution also initialization buffering area, and returns the notice success response of 506 to the SIPProxy/IMSCSCF230 with multi-path transmission session negotiation ability;

The SIPProxy/IMSCSCF230 508, with Multipath Transmission session negotiation ability returns the success response of multipath session establishment to calling party 210, and caller notification 210 liang of transmission paths distribution condition;

509, three transmission paths are had between calling party 210 and callee 220: a default path 260-D based on default route, the relay transmission path 260-R1 via media relay servers 250-1 and the relay transmission path 260-R2 via media relay servers 250-2.Use redundant transmission, the transmission means of media data packet as shown in Figure 4.

302, the receiving terminal of audio media data is arranged and initialization subflow buffering area, reassembly buffer district and jitter elimination window, receives the audio media data grouping that three transmission paths not exclusively intersected transmit.

In this step, described setting initialization subflow buffering area, reassembly buffer district and jitter elimination window, comprising:

Arrange the subflow buffering area with not exclusively crossing transmission path number same number, described subflow buffering area is for receiving the audio media data grouping of different transmission path, and each memory location of each described subflow buffering area of initialization is empty;

Reassembly buffer district is set, the sequence number of the audio media data grouping after described reassembly buffer district is used for storage redundancy rejecting, the sequence number storing the subflow buffering area of described audio media data grouping and described audio media data are grouped in the memory location in the subflow buffering area of storage, and described in initialization, reassembly buffer district size is N _jthe sequence number of the audio media data grouping after the redundant eliminating for storing in described reassembly buffer district operates all is initialized as-1, and the memory location that the sequence number of the subflow buffering area of described storage described audio media data grouping and described audio media data are grouped in the subflow buffering area of storage is all initialized as sky;

Arrange jitter elimination window, described jitter elimination window is for realizing the jitter elimination of described audio media data grouping, and the size of jitter elimination window described in initialization is W=100ms, W ∈ [W _min, W _max], W _min=40ms, W _max=160ms.

Particularly, when Fig. 6 shows the network implementation that second embodiment of the invention provides, the audio media data grouping rx-side buffering district of audio session media data multi-path redundancy transmission method arranges block diagram, as shown in Figure 6, Fig. 6 comprises subflow buffering area 610, reassembly buffer district 620 and jitter elimination window 630, during the network implementation that the second embodiment of the invention shown in composition graphs 7 provides, design drawing is implemented in the buffering area of the audio media data receiving terminal of audio session media data multi-path redundancy transmission method, Tu7Zhong reassembly buffer district is fixed, audio media data grouping is not exclusively intersected paths by many and is entered subflow buffering area 710 corresponding to each transmission path, wherein, the corresponding transmission path 260-D of subflow buffering area 710-D, the corresponding transmission path 260-R1 of subflow buffering area 710-R1, the corresponding transmission path 260-R2 of subflow buffering area 710-R2, reassembly buffer district 720 is entered by redundant eliminating operation, and then by jitter elimination window 730 size in dynamic adjustments reassembly buffer district 720, realize the jitter elimination of audio media data grouping, finally to audio media data grouping according to real-time audio session both sides consult audio coding decoding form and the network transmission protocol carry out readjustment decode, reduce original voice data.

The SIPProxy/IMSCSCF that Fig. 8 shows second embodiment of the invention to be provided participates in the encapsulation format schematic diagram of the audio media data grouping of the multi-path redundancy transmission of the multipath relay transmission service system of session negotiation, that form encapsulation adopts is Multipath Transmission agreement (MPTP), and in Fig. 8, the implication of each flag bit is as follows:

801: version number, 2bit, current is version 1;

802: type, 1bit, for illustration of the type (media data packet or control data grouping) of audio media data grouping;

803: filler, 1bit, indicate whether the padding data having non-effective load;

804: towards the MPTP type of special applications, 4bit, indicates the application-specific of such packet;

805: type of service, 4bit, indicate the transmission demand of different classs of business;

806: reserved field, 4bit, set to 0;

807: subflow sequence number, 16bit, for identification audio media data packet transmission sequence number in the path, in redundant transmission, it is identical with audio media data packet sequence number that subflow sequence number is set;

808: path identifier, 32bit, for identifying a transmission paths, deposit audio media data grouping for audio media data grouping receiving terminal by this;

809: sequence number, 32bit, for identifying audio media data grouping unique in real-time audio session, for audio media data grouping receiving terminal by this sequence number restructuring original audio data;

810: load, be real-time audio media data packet waiting for transmission.

In embody rule, subflow buffering area 610 is responsible for the audio media data received from many incomplete intersecting paths and is divided into groups, by carrying out redundant eliminating to audio media data grouping, and reassembly buffer district 620 is stored in the memory location that audio media data divides into groups, simultaneously according to the statistics of the audio media data grouping received and the rejecting situation of redundancy audio media data grouping, the polling priority of the subflow buffering area that the different transmission path of periodic adjustment subflow buffering area 610 is corresponding.By the size of dynamic adjustments jitter elimination window 630, can realize the jitter elimination of audio media data grouping, the readjustment decode procedure of audio media data grouping is simultaneously also by the size according to adjustment jitter elimination window during the fructufy of adjusting back.

303, the receiving terminal of audio media data carries out redundant eliminating to the described audio media data grouping received.

In this step, the redundant eliminating flow chart of the audio media data grouping of audio session media data multi-path redundancy transmission method when Fig. 9 shows the network implementation that second embodiment of the invention provides, concrete steps comprise:

901, initialization reassembly buffer district, size is N _j;

902, adopt the mode of poll from subflow buffering area, each path, obtain audio media data grouping i;

903, judge whether described audio media data grouping i pay(useful) load is empty, if so, then performs step 911, otherwise, perform step 904;

904, extract the sequence number of audio media data grouping i, sequence number 809 as shown in Figure 8, is designated as FSN _i;

905, m=FSN is calculated _imodN _j;

The sequence of the audio media data grouping of 906, in query reformulation buffering area, m position being deposited, same sequence number 809 as shown in Figure 8, is designated as J_FSN _m;

907, J_FSN is judged _mwhether equal-1, if so, then to illustrate that audio media data is grouped into the new reassembly buffer district that arrives, perform step 909, otherwise, perform step 908;

908, J_FSN is judged _mwhether be less than FSN _i, if so, then illustrate that audio media data grouping i is the grouping of newly arrived audio media data, perform step 909, otherwise, perform step 911;

909, by FSN _ivalue be assigned to J_FSN _m;

910, the memory location of i in subflow buffering area of being divided into groups by audio media data to be stored in reassembly buffer district in m position;

911, judge whether to terminate to receive, if so, terminate the reception of audio frequency media data packet, otherwise, perform step 902.

304, the receiving terminal of audio media data carries out reorganization operation to the described audio media data grouping after redundant eliminating operation, and the audio coding decoding form consulted according to real-time audio session both sides and network transmission protocol reduction original audio data.

In this step, described reorganization operation, comprising:

Dynamic adjustments jitter elimination window size;

According to described jitter elimination window size, carry out described audio media data grouping callback operation.

Further, described adjustment jitter elimination window size, comprising:

Record audio media data packet i arrives time R (i) in described reassembly buffer district, according to timestamp or other flag bits for record audio media data packet transmitting time in described audio media data grouping i, obtain the transmitting time S (i) of described audio media data grouping i, calculate delay variation J (i)=R (the i)-S (i) of described audio media data grouping i;

According to the audio media data grouping arriving described reassembly buffer district, obtain the expection delay variation of the audio media data grouping i newly arriving described reassembly buffer district wherein, N is fixed value, gets N=4, and i>=N, P (k) are weight coefficient, general value P (i-1)=0.5, P (i-2)=0.3, P (i-3)=0.125, P (i-4)=0.075;

Calculate the class standard error of described audio media data grouping i delay variation $J_{ARMSE}\left(i\right)=\sqrt{\frac{1}{N}-\underset{k=i-N+1}{\overset{i}{\&Sigma;}}{\&lsqb;J\left(k\right)-\stackrel{\&OverBar;}{J\left(k\right)}\&rsqb;}^2};$

Threshold value g is set ₁=5ms, g ₂=10ms is for judging jitter elimination window size change scope;

When increase jitter elimination window size, if J _aRMSE(i) ∈ [5,10], the increase of jitter elimination window if J _aRMSEi () > 10, jitter elimination window increases wherein, k=1;

When reduction jitter elimination window size, if J _aRMSE(i) ∈ [5,10], the reduction of jitter elimination window if J _aRMSEi () > 10, jitter elimination window reduces wherein, k=0.6;

If the size W of the jitter elimination window after regulating is less than jitter elimination window minimum W _min, then the size of jitter elimination window is set to jitter elimination window minimum W _minif the size W of the jitter elimination window after regulating is greater than jitter elimination window maximum W _max, then the size of jitter elimination window is set to jitter elimination window maximum W _max;

Work as J _aRMSE(i) < 5, the size of jitter elimination window is without the need to changing;

It should be noted that, in order to prevent present video media data packet i on the impact of the prediction accuracy of next audio media data grouping i+1, if revise jitter elimination window size, delay variation J (i) then changing audio media data grouping i is interval [J (i)-4, J (i)+4] in arbitrary integer, otherwise, do not change.

Further, described according to described jitter elimination window size, carry out described audio media data grouping callback operation, comprising:

The audio media data grouping that audio media data grouping serial number is FSN is found in jitter elimination window, the audio coding solution form consulted according to real-time audio session both sides and the network transmission protocol carry out readjustment decoding, search the audio media data grouping that audio media data grouping serial number is FSN+1 in jitter elimination window;

If find, be then the audio media data grouping readjustment decoding of FSN+1 by audio media data grouping serial number, continue to search the audio media data grouping that audio media data grouping serial number is FSN+2;

Otherwise, judge whether jitter elimination window has residue, if without residue, be then that the audio data packet of FSN+1 performs and once adjusts back decode operation by audio media data grouping serial number, continue to search the audio media data grouping that audio media data grouping serial number is FSN+2, otherwise, continue to search after waiting for preset time t, the desirable 10ms of t;

If continuous three times are searched successfully, then jitter elimination window is reduced the size of a Frame, judge now whether jitter elimination window is less than jitter elimination window minimum W _min, if so, then the size of jitter elimination window is set to jitter elimination window minimum W _min, continue to perform search operation;

If continuous three times are searched unsuccessful, then jitter elimination window is increased the size of a Frame, judge now whether jitter elimination window is greater than jitter elimination window maximum W _max, if so, then the size of jitter elimination window is set to jitter elimination window maximum W _max, continue to perform search operation.

In sum, during the network implementation of the present embodiment, audio session media data multi-path redundancy transmission method adopts multipath redundant transmission effectively can improve Discarded Packets compensation probability in transfer of data by conversation class voice data, ensures the reliability of transfer of data.And the reorganization operation of redundant eliminating and the jitter elimination based on dynamic buffering window is carried out by the receiving terminal divided into groups at audio media data, significantly reduce the impact that delay variation receives audio media data grouping, and then effectively promote customer service Quality of experience.

Last it is noted that above each embodiment is only in order to illustrate technical scheme of the present invention, be not intended to limit; Although with reference to foregoing embodiments to invention has been detailed description, those of ordinary skill in the art is to be understood that: it still can be modified to the technical scheme described in foregoing embodiments, or carries out equivalent replacement to wherein some or all of technical characteristic; And these amendments or replacement, do not make the essence of appropriate technical solution depart from the scope of the claims in the present invention.

Claims (10)

1. audio session media data multi-path redundancy transmission method during network implementation, is characterized in that, comprising:

The audio coding decoding form that the audio media data of catching is consulted according to real-time audio session both sides by the transmitting terminal of audio media data and network transmission protocol packing, adopt redundant fashion to be sent on many transmission paths not exclusively intersected simultaneously, carry out the transmission of real-time audio media data multi-path redundancy;

The receiving terminal of audio media data carries out redundant eliminating and reorganization operation to the grouping of the audio media data from different transmission path received, and the audio coding decoding form consulted according to real-time audio session both sides and network transmission protocol reduction original audio data.

2. audio session media data multi-path redundancy transmission method during network implementation according to claim 1, it is characterized in that, described many transmission paths not exclusively intersected comprise: a default path based on default route and one or more redundant transmission paths not exclusively intersected.

3. audio session media data multi-path redundancy transmission method during network implementation according to claim 1, it is characterized in that, arrange redundancy coefficient according to described real-time audio session, the number of described many transmission paths not exclusively intersected is greater than the redundancy coefficient of described real-time audio session setting.

4. audio session media data multi-path redundancy transmission method during network implementation according to claim 1, it is characterized in that, described redundant eliminating and reorganization operation, comprising:

Arrange and initialization subflow buffering area, reassembly buffer district and jitter elimination window;

The receiving terminal of described audio media data carries out redundant eliminating operation to the described audio media data grouping received;

The receiving terminal of described audio media data carries out reorganization operation to the described audio media data grouping after redundant eliminating operation.

5. audio session media data multi-path redundancy transmission method during network implementation according to claim 4, is characterized in that, described setting initialization subflow buffering area, reassembly buffer district and jitter elimination window, comprising:

Arrange the subflow buffering area with described not exclusively crossing transmission path number same number, described subflow buffering area is for receiving the audio media data grouping of different transmission path, and each memory location of each described subflow buffering area of initialization is empty;

Reassembly buffer district is set, the sequence number of the audio media data grouping after described reassembly buffer district is used for storage redundancy rejecting, the sequence number storing the subflow buffering area of described audio media data grouping and described audio media data are grouped in the memory location in the subflow buffering area of storage, and described in initialization, reassembly buffer district size is N _jthe sequence number of the audio media data grouping after the redundant eliminating for storing in described reassembly buffer district operates all is initialized as-1, and the memory location that the sequence number of the subflow buffering area of described storage described audio media data grouping and described audio media data are grouped in the subflow buffering area of storage is all initialized as sky;

Arrange jitter elimination window, described jitter elimination window is for realizing the jitter elimination of described audio media data grouping, and the size of jitter elimination window described in initialization is W, W ∈ [W _min, W _max].

6. audio session media data multi-path redundancy transmission method during network implementation according to claim 4, is characterized in that, described redundant eliminating operation, comprising:

The mode of S1, employing poll inquires about each described subflow buffering area, obtains the up-to-date described audio media data grouping i received, extracts the sequence number of described audio media data grouping i, be designated as FSN _i;

S2, by the sequence number FSN of described audio media data grouping i _iwith described reassembly buffer district size N _jcarry out modular arithmetic, obtain numerical value m, i.e. m=FSN _imodN _j, inquire about the sequence number of the audio media data grouping that m position in described reassembly buffer district stores, be designated as J_FSN _m;

If S3 is J_FSN _m=-1, then by FSN _ivalue be assigned to J_FSN _m, the sequence number of the subflow buffering area at described audio media data grouping i place and memory location are stored in m position, reassembly buffer district, perform step S1;

If S4 is J_FSN _m≠-1, FSN _i> J_FSN _m, then by FSN _ivalue be assigned to J_FSN _m, the sequence number of the subflow buffering area at described audio media data grouping i place and memory location are stored in m position, reassembly buffer district, perform step S1;

If S5 is J_FSN _m≠-1, FSN _i≤ J_FSN _m, then step S1 is performed.

7. audio session media data multi-path redundancy transmission method during network implementation according to claim 6, is characterized in that, the mode of described employing poll inquires about each described subflow buffering area, comprising:

The reception of the audio media data grouping in each described subflow buffering area of receiving terminal periodic statistical of described audio media data and redundant eliminating situation, obtain the polling priority sequence of subflow buffering area corresponding to different transmission path;

The receiving terminal of described audio media data, according to each described subflow buffering area of described polling priority sequence queries, carries out redundant eliminating operation.

8. audio session media data multi-path redundancy transmission method during network implementation according to claim 4, it is characterized in that, described reorganization operation, comprising:

Dynamic adjustments jitter elimination window size;

According to described jitter elimination window size, carry out described audio media data grouping callback operation.

9. audio session media data multi-path redundancy transmission method during network implementation according to claim 8, is characterized in that, described dynamic adjustments jitter elimination window size, comprising:

Record audio media data packet i arrives time R (i) in described reassembly buffer district, according to timestamp or other flag bits for record audio media data packet transmitting time in described audio media data grouping i, obtain the transmitting time S (i) of described audio media data grouping i, calculate delay variation J (i)=R (the i)-S (i) of described audio media data grouping i;

According to the audio media data grouping arriving described reassembly buffer district, obtain the expection delay variation of the audio media data grouping i newly arriving described reassembly buffer district wherein, N is fixed value, and i>=N, P (k) they are weight coefficient, and $\underset{k=i-N}{\overset{i-1}{\&Sigma;}}P\left(k\right)=1;$

Calculate the class standard error of described audio media data grouping i delay variation $J_{ARMSE}\left(i\right)=\sqrt{\frac{1}{N}\underset{k=i-N+1}{\overset{i}{\&Sigma;}}{\&lsqb;J\left(k\right)-\stackrel{\&OverBar;}{J\left(k\right)}\&rsqb;}^2};$

Threshold value g is set ₁, g ₂for judging jitter elimination window size change scope, and g ₁< g ₂;

When increase jitter elimination window size, if J _aRMSE(i) ∈ [g ₁, g ₂], jitter elimination window increases $2g_1+k|J\left(i\right)-\stackrel{\&OverBar;}{J\left(i\right)}|,$ If J _aRMSE(i) > g ₂, jitter elimination window increases $2g_2+\frac{k}{N}\underset{k=i-N+1}{\overset{i}{\&Sigma;}}|J\left(k\right)-\stackrel{\&OverBar;}{J\left(k\right)}|,$ Wherein, k ∈ [0.5,1];

When reduction jitter elimination window size, if J _aRMSE(i) ∈ [g ₁, g ₂], jitter elimination window reduces $2g_1+k|J\left(i\right)-\stackrel{\&OverBar;}{J\left(i\right)}|,$ If J _aRMSE(i) > g ₂, jitter elimination window reduces $2g_2+\frac{k}{N}\underset{k=i-N+1}{\overset{i}{\&Sigma;}}|J\left(k\right)-\stackrel{\&OverBar;}{J\left(k\right)}|,$ Wherein, k ∈ [0.5,1];

If the size W of the jitter elimination window after regulating is less than jitter elimination window minimum W _min, then the size of jitter elimination window is set to jitter elimination window minimum W _minif the size W of the jitter elimination window after regulating is greater than jitter elimination window maximum W _max, then the size of jitter elimination window is set to jitter elimination window maximum W _max;

Work as J _aRMSE(i) < g ₁, the size of jitter elimination window is without the need to changing;

If revise jitter elimination window size, then the expection delay variation of the audio media data grouping i changed for arbitrary integer in interval [J (i)-4, J (i)+4], otherwise, do not change.

10. audio session media data multi-path redundancy transmission method during network implementation according to claim 8, is characterized in that, described according to described jitter elimination window size, carries out described audio media data grouping callback operation, comprising:

The audio data packet that audio media data grouping serial number is FSN is found in jitter elimination window, the audio coding decoding form consulted according to real-time audio session both sides and the network transmission protocol carry out readjustment decoding, search the audio media data grouping that audio media data grouping serial number is FSN+1 in jitter elimination window;

If find, it is then the audio media data grouping readjustment decoding of FSN+1 by audio media data grouping serial number, continue to search the audio media data grouping that audio media data grouping serial number is FSN+2, otherwise, judge whether jitter elimination window has residue, if without residue, be then that the audio media data grouping of FSN+1 performs and once adjusts back decode operation by audio media data grouping serial number, continue to search the audio media data grouping that audio media data grouping serial number is FSN+2, otherwise, continue to search after waiting for preset time t;

If continuous three times are searched successfully, then jitter elimination window is reduced the size of a Frame, judge now whether jitter elimination window is less than jitter elimination window minimum W _min, if so, then the size of jitter elimination window is set to jitter elimination window minimum W _min, continue to perform search operation;

If continuous three times are searched unsuccessful, then jitter elimination window is increased the size of a Frame, judge now whether jitter elimination window is greater than jitter elimination window maximum W _max, if so, then the size of jitter elimination window is set to jitter elimination window maximum W _max, continue to perform search operation.