CN100550821C - Be used to eliminate the adaptive jitter buffering implementation method of ip voice data dithering - Google Patents
Be used to eliminate the adaptive jitter buffering implementation method of ip voice data dithering Download PDFInfo
- Publication number
- CN100550821C CN100550821C CNB031071961A CN03107196A CN100550821C CN 100550821 C CN100550821 C CN 100550821C CN B031071961 A CNB031071961 A CN B031071961A CN 03107196 A CN03107196 A CN 03107196A CN 100550821 C CN100550821 C CN 100550821C
- Authority
- CN
- China
- Prior art keywords
- packet voice
- formation
- voice frame
- pointer
- packet
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Abstract
The present invention relates to a kind of adaptive jitter buffering implementation method that is used to eliminate the ip voice data dithering, wherein adopt dithering cache with two-level pointer formation, detect the time jitter numerical value of each frame IP packet voice data earlier in real time, for the IP packet voice Frame of normal reception, directly send into second formation; If time jitter is excessive or packet loss arranged, then the pointer of this IP packet voice Frame is sent into first formation, in first formation, carry out order then and adjust, increase processing such as queue length or Discarded Packets compensation, and then be sent to second formation with the IP packet voice Frame pointer in first formation is disposable; If the real time shake always less than current processing delay, then reduces the length of first formation length of an IP packet voice Frame pointer in a period of time at every turn; And, corresponding IP packet voice Frame is at the uniform velocity outputed to next interface according to the IP packet voice Frame pointer of buffer memory in second formation.By above-mentioned steps, can adjust dynamically the processing delay of dithering cache according to the variation of time jitter, thereby reduce the adverse effect of time jitter, and can dynamically adjust according to packet drop.
Description
Technical field
The present invention relates to the IP technology, more particularly, relate to a kind ofly be used for eliminating ' the adaptive jitter buffering implementation method of speech data shake.
Background technology
Because the transmission characteristic of IP network, can cause to a certain degree time jitter to the packet voice data.Time-delay is the key factor that directly has influence on the ip voice quality, also is the basic reason that produces shake, and the time-delay in IP network mainly comprises propagation delay and handles time-delay.Propagation delay is meant caused time-delay when electric current or photon transmit a segment distance with a certain speed in network; It then is caused by the equipment of various processed voice in the voice transfer process handling time-delay, the router that for example uses in the internet, LAN (local area network (LAN)) switch, WAN (wide area network) switch etc.The time-delay that changes will produce shake, promptly because the variation of various time-delays will cause IP packet voice Frame to change through the speed that network arrives receiving terminal.The adverse effect that shake brings is the serious distortion that possible cause voice quality.
Therefore, must carry out the data caching to eliminate time jitter at data receiver, add buffering area by receiving terminal at speech processing device, be used to preserve the data of sufficiently long time, the slowest packet also can be arrived before voice restoration is handled timely, thereby eliminate the adverse effect of shake.Data in buffer quantity is many more, and the data dithering degree that receiving terminal can bear is just big more, and the packet loss that receives data is just more little.Packet loss is meant the ratio of the shared total packet of packet that the packet voice packet loses in the transmission course of IP network.The packet voice packet normally adopts UDP (user datagram protocol) agreement to transmit, because UDP is a kind of non-reliable, connectionless datagram service, itself does not guarantee service quality, and the resources such as bandwidth of internet fluctuate greatlyyer with the variation of number of users and traffic carrying capacity, and these all are the factor that produces packet loss for for example congested and network failures.Packet loss can influence the quality that voice restore, and after packet loss surpasses to a certain degree, can cause voice to restore.
Can produce time-delay but data are carried out buffer memory, data in buffer quantity is many more, and the time-delay of data is also just big more.Because the quality the when time-delay of speech data can influence the speech data playback too is so the time-delay of data and packet loss can not reach the most desirable index simultaneously.Dithering cache is commonly referred to JitterBuffer, and Fig. 1 is under the particular network resource situation, concerns schematic diagram between JitterBuffer processing delay and the corresponding packet loss.
As can be seen from Figure 1, processing delay hour, packet loss is very big; Packet loss hour, then processing delay is very big again.In order to take into account processing delay and packet loss, preferably can carry out real-time dynamic adjustment to JitterBuffer at the time jitter situation of actual reception data.Yet the scheme of also JitterBuffer dynamically not being adjusted in the prior art fails to address the above problem always.
Summary of the invention
At the above-mentioned defective of prior art, the present invention will carry out dynamic, real-time adjustment to the processing delay of dithering cache, is keeping under the prerequisite of voice quality, realizes minimizing of voice time-delay.
The technical solution adopted for the present invention to solve the technical problems is: a kind of adaptive jitter buffering implementation method that is used to eliminate IP packet voice data dithering is provided, employing has the dithering cache of two-level pointer formation, from IP network, receive frame IP packet voice data at every turn, and dynamically adjust according to the following steps:
(1) from IP network, receives frame IP packet voice data,, detect and preserve current time jitter numerical value in real time, detect again whether IP packet voice Frame pointer is arranged in current first formation according to the entrained parameter of this frame IP packet voice data;
(2), judge again then whether a current IP packet voice Frame and a last IP packet voice Frame be continuous if there is not IP packet voice Frame pointer in current first formation;
If (2-1) discontinuous, then current IP packet voice Frame pointer sent into first formation, and in first formation, reserve IP packet voice Frame pointer space for the pointer that should receive temporary transient unreceived IP packet voice Frame;
If (2-2) continuous, then directly current IP packet voice Frame pointer sent into second formation, and in first formation, reserve IP packet voice Frame pointer space for the pointer of this IP packet voice Frame;
(3) if IP packet voice Frame pointer is arranged in current first formation, then detect according to the IP packet voice Frame pointer space of existing IP packet voice Frame pointer and reservation in first formation whether it full again,
If (3-1) first formation less than, then directly forward step (3-3) to;
If (3-2) first formation is full, whether the current processing delay that then detects dithering cache has reached maximum;
If (3-2-1) do not reach maximum, then the length of first and second formations is increased the length of an IP packet voice Frame pointer respectively, again in the IP packet voice Frame pointer space in first formation that reserves in the pointer inserting step (2-1) with a compensation IP packet voice Frame, and the pointer of described compensation IP packet voice Frame sent to the latest position of second formation, enter step (3-3);
If (3-2-2) reached maximum, the compensation IP packet voice Frame that then inserts right quantity makes the IP packet voice Frame of the IP packet voice Frame pointer correspondence in first formation continuous, and the pointer of insert compensation IP packet voice Frame sent in first formation to receiving the position of the pointer space that temporary transient unreceived IP packet voice Frame is reserved, be sent to second formation with all the IP packet voice Frame pointers in first formation are disposable then, again current IP packet voice Frame pointer is sent into second formation, skips steps (3-3) is carried out subsequent step then;
(3-3) current IP packet voice Frame pointer is sent into first formation, judge again in first formation the pairing IP packet voice of IP packet voice Frame pointer Frame whether with second formation in the pairing IP packet voice of IP packet voice Frame pointer Frame continuous, if, then be sent to second formation with the IP packet voice Frame pointer in first formation is disposable, otherwise wait for next IP packet voice Frame;
(4) each frame IP packet voice data is carried out above-mentioned institute in steps, and, corresponding IP packet voice Frame is at the uniform velocity outputed to next interface according to the IP packet voice Frame pointer of buffer memory in second formation.
In method of the present invention, in described step (4) before, further comprising the steps of:
(A) the measured current time jitter numerical value and the current processing delay of dithering cache are compared, judge current time jitter numerical value whether in the time of one section pre-fixed length always less than the current processing delay of dithering cache;
(B) if then from second formation, delete an IP packet voice Frame pointer, and first formation reduced the length of an IP packet voice Frame pointer.
In method of the present invention, in described step (B), the operation that the IP packet voice Frame pointer in second formation is deleted is carried out the pointer of non-transmission IP packet voice Frame in only during quiet.
In method of the present invention, in described step (B), whether the current processing delay that will detect dithering cache has earlier reached minimum value, only when not reaching minimum value, just from second formation, delete an IP packet voice Frame pointer, and the length of first formation is reduced the length of an IP packet voice Frame pointer.
In method of the present invention, the time of described pre-fixed length is the total time length of 100 IP packet voice Frames.
In method of the present invention, in described step (3-2-2), the compensation IP packet voice Frame that is inserted is used to replace wherein discontinuous position to be leaked the IP packet voice Frame of receipts, and each compensation IP packet voice Frame is the result who carries out the linear prediction gained according to its former frame IP packet voice data.
In method of the present invention, in described step (3-2-2), if desired the compensation IP packet voice Frame of Cha Ruing total time length greater than 60ms, then inserting earlier total time length is the compensation IP packet voice Frame of 60ms, remainder inserts comfort noise.
In method of the present invention, the processing delay initial value of described dithering cache is the time span of 3 IP packet voice Frames.
Pass through such scheme, can be according to the variation of the time jitter of each IP packet voice Frame or the generation of packet drop, processing delay to dithering cache carries out dynamic self adaptation adjustment, the processing time-delay of dithering cache is rationalized, thereby reduce the adverse effect of the shake of IP network speech data the speech data playback quality.In addition, the detection of packet loss and processing operation all have been embedded in the processing procedure of dithering cache, can further improve the playback quality of speech data.
Description of drawings
The invention will be further described below in conjunction with drawings and Examples, in the accompanying drawing:
Fig. 1 is under the particular network resource situation, concerns schematic diagram between the processing delay of dithering cache and the corresponding packet loss;
Fig. 2 is a workflow diagram of adjusting the processing delay of dithering cache among the present invention with the two-level pointer formation;
Fig. 3 is the schematic diagram when IP packet voice Frame pointer normally transmits among the present invention;
Fig. 4 is a schematic diagram of adjusting IP packet voice data frame sequence among the present invention in first formation;
Fig. 5 is the schematic diagram when increasing the length of first, second formation among the present invention.
Embodiment
One, the data structure of JitterBuffer
Among the present invention, the data structure of JitterBuffer adopts the two-stage formation to realize.The structure of formation is a pointer alignment, be cached with the current address pointer of IP packet voice data in each grade formation, the purpose that adopts pointer alignment is only to need transmission pointer rather than each step all to transmit IP packet voice data in the process of data carrying, thereby can avoid the copy of data.The purpose that adopts the two-stage formation is in order to simplify the complexity of JitterBuffer operation control, separate management to be carried out in the output and the input operation of data.The input operation of data is responsible in first formation wherein, can achieve a butt joint to receive order adjustment, the Discarded Packets compensation of data; The output function of data is responsible in second formation, is the outbound data output interface that JitterBuffer handles, and can realize that the time interval between the packet is adjusted, and guarantees the at the uniform velocity output of data.The maximum length of first formation and second formation is the maximum processing delay of JitterBuffer just.
Two, the dynamic adjustment of JitterBuffer processing delay
In the inventive method, the basic procedure that processing delay is dynamically adjusted as shown in Figure 2.Wherein the initial treatment time delay of JitterBuffer can be set to the time span of 3 IP packet voice Frames, is example with coded data G.729, and then its initial treatment time delay is 30ms;
Dithering cache of the present invention will detect and preserve the time jitter of its each IP packet voice Frame that receives in real time when receiving IP packet voice Frame.
1, the normal reception handled
As can be seen from Figure 2, the flow process when normally receiving is a step 1,2,3,4,5,17 and step 21, wherein:
To step 4, do not have IP packet voice Frame pointer in first formation this moment by step 3, the reception of expression front is normal always;
To step 5, a current IP packet voice Frame and a last IP packet voice Frame are continuous by step 4, and expression does not have packet loss;
To step 21, illustrate that the real time shake is normal, not always less than the current processing delay of dithering cache, does not need to reduce the length of formation by step 17 in the time of one section pre-fixed length.
As shown in Figure 3, when normally receiving, all IP packet voice Frame pointers can be skipped first formation, directly deliver to second formation, at the uniform velocity output to next interface by second formation again.
2, actual jitter is big or packet loss arranged
As can be seen from Figure 2, by step 4 to step 6, explanation is owing to the reason of shake than big or packet loss, received and the discontinuous IP packet voice of last IP packet voice Frame Frame, one or more IP packet voice Frames of its front have been received in leakage, and send this IP packet voice Frame pointer into first formation this moment, when receiving the next frame data, first formation no longer is empty, will jump to step 7 by step 3.Shown in Fig. 4,5 Far Left, the 1st IP packet voice Frame is normal the reception, its pointer is put in second formation, " 1 " shown in Fig. 4,5 Far Left secondary series, simultaneously in first formation, reserve IP packet voice Frame pointer space, as " 1 ' " shown in Fig. 4,5 Far Left first are listed as for this IP packet voice Frame; Should receive the 2nd IP packet voice Frame afterwards, but received the 3rd IP packet voice Frame, so the 3rd IP packet voice Frame pointer is put in first formation, " 3 " shown in Fig. 4,5 Far Left, first row, simultaneously for should receiving that temporary transient unreceived IP packet voice Frame (the 2nd IP packet voice Frame) reserves IP packet voice Frame pointer space, " 2 ' " shown in Fig. 4,5 Far Left first row.
3, first formation less than but do not need to increase queue length
As can be seen from Figure 2, Ci Shi flow process is to step 13 by step 7,11,12.
By step 7 to step 11, the reason of expression or packet loss big owing to the shake of front, have at least a pointer to be placed in first formation with front discontinuous IP packet voice Frame, and first formation less than, this moment current IP packet voice Frame pointer is also put into first formation.Shown in Fig. 4,5 second row, when receiving the 4th IP packet voice Frame, its pointer also is put in first formation, as Fig. 4,5 second ranked first " 4 " shown in the row.
By step 12 to step 13, represent the IP packet voice Frame that current IP packet voice Frame pointer just in time corresponding front leakage of putting into first formation is received, make each IP packet voice Frame pointer in first formation become continuous state, be sent to second formation with all the IP packet voice Frame pointers in first formation are disposable this moment.Shown in the 3rd, the 4th row of Fig. 4, after the 4th IP packet voice Frame, received the 2nd IP packet voice Frame again, it before is the position that the 2nd IP packet voice Frame pointer reserved that its pointer is placed on, and ranked first " 2 " shown in the row as Fig. 4 the 3rd.At this moment the 2nd IP packet voice Frame and the 3rd, the 4th constitutes three continuous IP packet voice Frames, simultaneously continuous with the 1st IP packet voice Frame of second formation, so just can be sent to second formation with these three IP packet voice Frame pointers are disposable.
4, full and its length of first formation does not reach maximum
As can be seen from Figure 2, Ci Shi flow process is to step 11 by step 7,8,9,10.
To step 8, expression have at least a pointer with front discontinuous IP packet voice Frame to be placed in first formation, and first formation is full because the shake of front is big or packet loss by step 7.Shown in second row of Fig. 5, queue length originally is the length of 4 IP packet voice Frame pointers, and when receiving the 5th IP packet voice Frame, first formation is full.
By step 8 to step 9,10,11, though represent that first formation is full, but its length does not reach maximum (being 200ms in the present embodiment), shown in the 3rd row of Fig. 5, need the length of first, second formation is increased the length of an IP packet voice Frame pointer this moment, queue length becomes the length of 5 IP packet voice Frame pointers, and then inserts a compensation IP packet voice Frame A and its pointer is sent to second formation, ranked second " A " shown in the row as Fig. 5 the 3rd.Again the 5th IP packet voice Frame pointer sent into first formation, ranked first " 5 " shown in the row as Fig. 5 the 3rd.
Wherein, owing to do not receive the 2nd IP packet voice Frame, so the compensation IP packet voice Frame A that inserts is used to replace the 2nd IP packet voice Frame, this compensation IP packet voice Frame A is the result who carries out the linear prediction gained according to the 1st frame data.When if a plurality of IP packet voice Frame packet loss is arranged, for fear of in the process of compensation, producing dumb sound, insert compensation IP packet voice Frame total time length should be greater than 60ms, for JitterBuffer increment greater than 60ms, inserting after total time, length was the compensation IP packet voice Frame of 60ms, remainder will insert comfort noise.
In said process, the increase of the processing delay of JitterBuffer is real-time, thereby can handle at the data dithering situation that the IP network pleurapophysis is sent out, to eliminate its adverse effect.
5, full and its length of first formation has reached maximum, can not increase queue length
To step 14,15,16, represent that first formation is full by step 8, and its length reached maximum, can not increase queue length this moment again.The compensation IP packet voice Frame that should insert right quantity this moment makes the pairing IP packet voice of the IP packet voice Frame pointer Frame in first formation continuous, and the pointer of insert compensation IP packet voice Frame is sent to the position for receiving that temporary transient unreceived IP packet voice Frame pointer is reserved in first formation.Suppose not receive the 2nd IP packet voice Frame, then a pointer that compensates IP packet voice Frame is inserted in the position of the 2nd IP packet voice Frame pointer in first formation, be sent to second formation with all the IP packet voice Frame pointers in first formation are disposable then, again current IP packet voice Frame pointer sent into second formation.
6 reduce the processing delay of JitterBuffer
As can be seen from Figure 2, among the present invention, whenever receive an IP packet voice Frame, all to compare with the current processing delay of measured time jitter and JitterBuffer, simultaneously with reference to RTCP (Realtime Transport Control Protocol, RTCP Real-time Transport Control Protocol) the calculation of parameter result of data, if the shake of finding network in the time period (for example total time length of 100 IP packet voice Frames) of pre-fixed length always less than the current processing delay of JitterBuffer, then need to reduce the processing delay of JitterBuffer, concrete mode is to delete an IP packet voice Frame pointer earlier from second formation, the length of first formation is reduced the length of an IP packet voice Frame pointer again.
In order to guarantee certain queue length, whether the current processing delay that will detect dithering cache has earlier reached minimum value (being made as 30ms in the present embodiment), the i.e. time span of 3 IP packet voice Frames, only when its length does not reach minimum value, just from second formation, delete an IP packet voice Frame pointer, and the length of first formation is reduced the length of an IP packet voice Frame pointer.
Concrete reduce to operate should be during quiet the pointer of the non-transmission IP packet voice Frame correspondence in first formation to be deleted realize, the purpose of doing like this is for fear of the normal speech data of deletion voice quality to be caused extra influence, thereby can guarantee not influence normal speech data.
Among the present invention, all will carry out flow process shown in Figure 2, and carry out corresponding step by its judged result to each IP packet voice Frame.Wherein the IP packet voice Frame pointer transmission between first formation and second formation is real-time, in the no time interval, pass to second formation so promptly IP packet voice Frame pointer that at once will dispose after first queue processing finishes is disposable.Therefore, though the IP packet voice Frame pointer that receives when dynamically adjusting will be through the processing of two-stage formation, the overall time-delay of IP packet voice Frame pointer is the time-delay of a formation normally the time.
Claims (8)
1, a kind of adaptive jitter buffering implementation method that is used to eliminate IP packet voice data dithering, it is characterized in that, employing has the dithering cache of two-level pointer formation, receives frame IP packet voice data at every turn from IP network, and dynamically adjusts according to the following steps:
(1) from IP network, receives frame IP packet voice data,, detect and preserve current time jitter numerical value in real time, detect again whether IP packet voice Frame pointer is arranged in current first formation according to the entrained parameter of this frame IP packet voice data;
(2), judge again then whether a current IP packet voice Frame and a last IP packet voice Frame be continuous if there is not IP packet voice Frame pointer in current first formation;
If (2-1) discontinuous, then current IP packet voice Frame pointer sent into first formation, and in first formation, reserve IP packet voice Frame pointer space for the pointer that should receive temporary transient unreceived IP packet voice Frame;
If (2-2) continuous, then directly current IP packet voice Frame pointer sent into second formation, and in first formation, reserve IP packet voice Frame pointer space for the pointer of this IP packet voice Frame;
(3) if IP packet voice Frame pointer is arranged in current first formation, then detect according to the IP packet voice Frame pointer space of existing IP packet voice Frame pointer and reservation in first formation whether it full again,
If (3-1) first formation less than, then directly forward step (3-3) to;
If (3-2) first formation is full, whether the current processing delay that then detects dithering cache has reached maximum;
If (3-2-1) do not reach maximum, then the length of first and second formations is increased the length of an IP packet voice Frame pointer respectively, again in the IP packet voice Frame pointer space in first formation that reserves in the pointer inserting step (2-1) with a compensation IP packet voice Frame, and the pointer of described compensation IP packet voice Frame sent to the latest position of second formation, enter step (3-3);
If (3-2-2) reached maximum, the compensation IP packet voice Frame that then inserts right quantity makes the IP packet voice Frame of the IP packet voice Frame pointer correspondence in first formation continuous, and the pointer of insert compensation IP packet voice Frame sent in first formation to receiving the position of the pointer space that temporary transient unreceived IP packet voice Frame is reserved, be sent to second formation with all the IP packet voice Frame pointers in first formation are disposable then, again current IP packet voice Frame pointer is sent into second formation, skips steps (3-3) is carried out subsequent step then;
(3-3) current IP packet voice Frame pointer is sent into first formation, judge again in first formation the pairing IP packet voice of IP packet voice Frame pointer Frame whether with second formation in the pairing IP packet voice of IP packet voice Frame pointer Frame continuous, if, then be sent to second formation with the IP packet voice Frame pointer in first formation is disposable, otherwise wait for next IP packet voice Frame;
(4) each frame IP packet voice data is carried out above-mentioned institute in steps, and, corresponding IP packet voice Frame is at the uniform velocity outputed to next interface according to the IP packet voice Frame pointer of buffer memory in second formation.
2, method according to claim 1 is characterized in that, and is in described step (4) before, further comprising the steps of:
(A) the measured current time jitter numerical value and the current processing delay of dithering cache are compared, judge current time jitter numerical value whether in the time of one section pre-fixed length always less than the current processing delay of dithering cache;
(B) if then from second formation, delete an IP packet voice Frame pointer, and the length of first formation reduced the length of an IP packet voice Frame pointer.
3, method according to claim 2 is characterized in that, in described step (B), the operation that the IP packet voice Frame pointer in second formation is deleted is carried out the pointer of non-transmission IP packet voice Frame in only during quiet.
4, method according to claim 3, it is characterized in that, in described step (B), whether the current processing delay that will detect dithering cache has earlier reached minimum value, only when not reaching minimum value, just from second formation, delete an IP packet voice Frame pointer, and the length of first formation is reduced the length of an IP packet voice Frame pointer.
According to each described method among the claim 2-4, it is characterized in that 5, the time of described pre-fixed length is the total time length of 100 IP packet voice Frames.
6, according to each described method among the claim 1-4, it is characterized in that, in described step (3-2-2), the compensation IP packet voice Frame that is inserted is used to replace wherein discontinuous position to be leaked the IP packet voice Frame of receipts, and each compensation IP packet voice Frame is the result who carries out the linear prediction gained according to its former frame IP packet voice data.
7, method according to claim 6, it is characterized in that, in described step (3-2-2), if desired the compensation IP packet voice Frame of Cha Ruing total time length greater than 60ms, then insertion total time length is the compensation IP packet voice Frame of 60ms earlier, and remainder inserts comfort noise.
According to each described method among the claim 1-4, it is characterized in that 8, the processing delay initial value of described dithering cache is the time span of 3 IP packet voice Frames.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031071961A CN100550821C (en) | 2003-03-13 | 2003-03-13 | Be used to eliminate the adaptive jitter buffering implementation method of ip voice data dithering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031071961A CN100550821C (en) | 2003-03-13 | 2003-03-13 | Be used to eliminate the adaptive jitter buffering implementation method of ip voice data dithering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1531276A CN1531276A (en) | 2004-09-22 |
| CN100550821C true CN100550821C (en) | 2009-10-14 |
Family
ID=34282913
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB031071961A Expired - Fee Related CN100550821C (en) | 2003-03-13 | 2003-03-13 | Be used to eliminate the adaptive jitter buffering implementation method of ip voice data dithering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100550821C (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1791055B (en) * | 2004-12-14 | 2010-04-07 | 腾讯科技（深圳）有限公司 | Method for decreasing data packet broadcast time-delay |
| CN100433728C (en) * | 2006-07-19 | 2008-11-12 | 华为技术有限公司 | Speech data frame receiving method and terminal, wireless packet network gate and system |
| CN101119323A (en) * | 2007-09-21 | 2008-02-06 | 腾讯科技（深圳）有限公司 | Method and device for solving network jitter |
| CN101222288B (en) * | 2008-02-01 | 2011-07-20 | 华为技术有限公司 | IP network transmission method, system and equipment automatically adapting network jitter |
| CN101335796B (en) * | 2008-07-25 | 2011-01-19 | 华为技术有限公司 | Decoding scheduling method and device |
| CN101369880B (en) * | 2008-09-28 | 2012-09-19 | 华为技术有限公司 | Detection processing method and apparatus for time label jumping |
| CN101771599B (en) * | 2008-12-26 | 2012-07-04 | 中国移动通信集团公司 | Method and device for receiving RTP data packets |
| JP5369814B2 (en) * | 2009-03-26 | 2013-12-18 | ソニー株式会社 | Receiving device and time correction method for receiving device |
| CN101742333A (en) * | 2009-12-17 | 2010-06-16 | 中兴通讯股份有限公司 | Stable output method of media data and device |
| EP2770504B1 (en) * | 2011-10-20 | 2016-11-30 | LG Electronics Inc. | Method of managing a jitter buffer, and jitter buffer using same |
| CN104853231B (en) | 2014-02-19 | 2019-06-14 | 腾讯科技（北京）有限公司 | Information processing method, equipment and system |
| CN104240716B (en) * | 2014-06-11 | 2017-04-26 | 杭州联汇数字科技有限公司 | Audio data quality optimization method |
| CN106209496A (en) * | 2016-06-23 | 2016-12-07 | 珠海格力电器股份有限公司 | A kind of communication abnormality detection device and method |
| CN109413485B (en) * | 2017-08-17 | 2022-02-01 | 成都鼎桥通信技术有限公司 | Data caching method and device |
| CN114513477A (en) * | 2020-11-17 | 2022-05-17 | 华为技术有限公司 | Message processing method and related device |
-
2003
- 2003-03-13 CN CNB031071961A patent/CN100550821C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN1531276A (en) | 2004-09-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100550821C (en) | Be used to eliminate the adaptive jitter buffering implementation method of ip voice data dithering | |
| US7079486B2 (en) | Adaptive threshold based jitter buffer management for packetized data | |
| EP1535419B1 (en) | Method and devices for controlling retransmissions in data streaming | |
| JP3882187B2 (en) | Flow control system and method | |
| US6452950B1 (en) | Adaptive jitter buffering | |
| CN1754335B (en) | Flow control in a packet data communication system | |
| US6952407B2 (en) | Minimizing latency with content-based adaptive buffering | |
| US7102998B1 (en) | Scaleable congestion control method for multicast communications over a data network | |
| KR20010074883A (en) | System and method for low latency communication | |
| CN100502349C (en) | Method for increasing transmission rate of data packet initiated by receiver | |
| US20040071085A1 (en) | System and method for a transmission rate controller | |
| CN101523822B (en) | Voice transmission apparatus | |
| US7411978B2 (en) | Method of transmitting data packets using RTP and RTCP protocols | |
| US7787500B2 (en) | Packet receiving method and device | |
| JP2001160824A (en) | Wired and wireless mixed network data distributor and data distribution method | |
| EP2380294B1 (en) | Method for controlling a flow in a packet switching network | |
| CN103905328A (en) | Data transmission control system and method and related equipment | |
| Liu et al. | Improving explicit congestion notification with the mark-front strategy | |
| EP1471695B1 (en) | Method for flow control in a communication system | |
| US20030210681A1 (en) | Method and apparatus for controlling delay fluctuation absorption buffer | |
| WO2008068167A1 (en) | Method of adaptively dejittering packetized signals buffered at the receiver of a communication network node | |
| JP2007221290A (en) | Device and system for transmitting packet | |
| JP2006050488A (en) | Communication terminal device and buffer control method | |
| CN1281036C (en) | Packet switching network distributed adaptive dither buffer adjusting method | |
| WO2023155747A1 (en) | Data coding bitrate adaptive adjustment method, apparatus and device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20091014 Termination date: 20130313 |