CN100341267C - Method for realizing real-time data stream shaking reduction of mixed packaging time length - Google Patents
Method for realizing real-time data stream shaking reduction of mixed packaging time length Download PDFInfo
- Publication number
- CN100341267C CN100341267C CNB2004101009048A CN200410100904A CN100341267C CN 100341267 C CN100341267 C CN 100341267C CN B2004101009048 A CNB2004101009048 A CN B2004101009048A CN 200410100904 A CN200410100904 A CN 200410100904A CN 100341267 C CN100341267 C CN 100341267C
- Authority
- CN
- China
- Prior art keywords
- passage
- time length
- bunch
- packaging time
- channel
- 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
Landscapes
- Time-Division Multiplex Systems (AREA)
Abstract
The present invention discloses a method for realizing the buffeting eliminating of the real-time data stream of the mixed packaging time length. The present invention solves the problem that the prior art can not realize the buffer storage, the buffeting elimination and the disorder elimination of the real-time data stream of the mixed packaging time length. The present invention adopts different channel occupying table to identify the channel occupying situation of different packaging time length, when the start data frame of the packaging time length is received, a corresponding channel occupying table is looked up to distribute a storage channel and a storage data frame for the current packaging time length, and the storage channel of each channel occupying table is marked with occupation. When the data packets are read, the channel information is read orderly by using the least time length as the time interval, and the corresponding channel information is read according to the packaging time length in the channel information and the storage location.
Description
Technical field
The present invention relates to the real time data stream transmission technology, relate in particular to a kind of implementation method of real-time data stream shaking reduction of mixed packaging time length.
Background technology
On IP network during transmitting realtime stream, because the uncertainty of network delay, can introduce the time jitter of real-time stream and out of order, in order to guarantee the quality of stream in real time, need carry out buffer memory at receiving terminal and go to tremble, eliminate out of order.Under the different application situation, terminal has different encoding and decoding speed, causes various real-time streams when the receiving terminal buffer memory goes to tremble, and need carry out buffer memory according to the packaging time length that difference flows in real time and go to tremble, to eliminate shake.
Number of patent application is 02119584.6, and the technical scheme that name is called the Chinese patent application of " a kind of buffer memory of realizing removes the large capacity realtime stream processing method of trembling " has adopted the method for queue scheduling to realize that the buffer memory of stream goes to tremble in real time.As shown in Figure 1, buffer memory goes to tremble the queue scheduling flow process and is: at first, input receives the real-time streaming data bag, and the real-time streaming data bag is deposited in input-buffer; If have complete real-time streaming data bag in the input-buffer, then the request of writing real-time stream packets is sent in the arbitration of the read-write in queue scheduling, and provides the channel number of writing real-time stream packets of request simultaneously; Secondly, transmitting time in the queue scheduling safeguards that the flow process unification safeguards the transmitting time of real-time streaming data bag in all passages, and at interval send the request of reading the real-time streaming data bag with regular time to the read-write arbitration, provide the channel number of reading the real-time streaming data bag of request and real-time streaming data simultaneously and wrap in memory address in the shared buffer memory; Once more, read-write arbitration is arbitrated reading request signal and written request signal, and starts and write real-time stream packets flow process, and perhaps real-time streaming data bag flow process is read in startup; Write real-time stream packets flow process and read the affiliated channel information of the real-time stream packets of writing of request, calculate and provide and write the memory address of real-time stream packets in shared buffer memory, and these real-time streaming data bag data are written to the shared buffer memory from input-buffer; Read the channel information under the real-time streaming data bag read that real-time streaming data bag flow process reads request, provide this and read real-time streaming data and wrap in address in the shared buffer memory, from shared buffer memory, read real-time streaming data bag data then, deposit output buffers in; At last, the outgoing inspection output buffers if find that complete real-time streaming data bag is arranged, then sends the real-time streaming data bag.
Though said method can go to tremble in the receiving terminal realization of stream in real time, but the mode of the transmitting time of this employing fixed intervals can only be handled fixing single packaging time length, under mixed packaging time length (promptly comprising two different packaging time lengths at least) situation, fixedly the packaging time length buffer memory that can not solve all real-time stream goes to tremble, eliminate out of order problem.The mixing voice stream of suppose there are 5 milliseconds, 20 milliseconds two kinds being packed at interval carries out buffer memory according to the method for an embodiment who describes in the above-mentioned patent application and goes to tremble, eliminates out of order will going wrong.Because, in writing the voice packet flow process, the position that voice packet is write in calculating all is to handle with the Fixed Time Interval of 20ms, calculate the packaging time length voice packet at interval of 5ms with the time interval algorithm of 20ms, can cause the writing position that calculates incorrect, thereby packet loss take place, coat situations such as lid.Equally, in reading the voice packet flow process, send the voice packet of the packaging time length of 5ms, can lose nearly 3/4 voice packet, cause voice unavailable according to fixing 20ms transmission time interval.
Summary of the invention
The invention provides a kind of implementation method of real-time data stream shaking reduction of mixed packaging time length, can't go to tremble and eliminate out of order problem to the real time data stream buffer memory of mixed packaging time length to solve prior art.
A kind of implementation method of real-time data stream shaking reduction of mixed packaging time length; This method comprises:
Safeguard step, be used to be respectively the passage that each packaging time length safeguards that reflection passage takies situation and take table;
Allocation step is used for when receiving the beginning Frame of a packing duration, searches corresponding passage and takies table and come distribution of flows memory channel for this packaging time length, and take in the table this memory channel is labeled as at all passages and take;
Write step is used to calculate the Frame that receives and should writes the position of memory channel, and is written in this memory channel;
Reading step is used for time interval of minimum packaging time length time slot as data dispatching order poll passage and sense data frame from memory channel.
Wherein:
Concern according to the multiple between all the other each packaging time lengths and the maximum packaging time length except that maximum packaging time length, the passage of packaging time length correspondence is taken passage in the table be divided into passage bunch, make the port number of each passage bunch correspondence equal described multiple, and each passage in each passage bunch take a time slot.
In safeguarding step, go back maintenance channel bunch state table, the corresponding relation between time slot that the memory channel of this state table record packaging time length, this packaging time length takies and the passage bunch number; Time slot that memory channel is shared, corresponding passage bunch number and corresponding packaging time length are recorded in this passage bunch state table in described allocation step.
In the described allocation step, will take at the passage of current packaging time length correspondence in the table to take all to be labeled as in the table at each passage in the rest channels in the same passage bunch and take, this rest channels can not be reallocated with the memory channel that distributes.
In the described allocation step, when other packaging time length memory allocated passages of giving except that hand-to-hand fight in tights bag duration, to take at the passage of hand-to-hand fight in tights bag duration in the table with the memory channel that distributes in same passage bunch and take in the table and the described memory channel passage in same passage bunch not at other passages, take other passages the table at passage and take all to be labeled as in the table and take, make these passages no longer distribute to other packaging time lengths except that current packaging time length correspondence.
Reading step specifically comprises the steps:
Number inquire about described passage bunch state table according to the passage of the current channel number correspondence of poll bunch and obtain corresponding packaging time length;
Judge whether current packaging time length is maximum packaging time length, if, the pairing passage sense data of current channel number frame then from being polled to, otherwise, from passage bunch state table, obtain time slot corresponding and number form channel number, and from the pairing passage sense data of the channel number of this composition frame.
The beneficial effect that the present invention brings mainly contains:
1, the present invention is by taking table for the real time data stream maintenance channel of different packaging time lengths, and take passage in the table according to the passage of packaging time length correspondence and take multiple between situation and current packaging time length and other packaging time lengths and concern and distribute passage, thereby the buffer memory that can solve the real-time stream of mixed packaging time length goes to tremble, eliminate out of order problem.
2, apply the present invention on the network equipment, can well satisfy the demand of the wireless and fixed network fusion of product, realize wireless and integrated compatible equipment fixed network.
Description of drawings
Fig. 1 goes to tremble the queue scheduling FB(flow block) for buffer memory in the prior art;
Fig. 2 takies the corresponding relation schematic diagram of passage in the table bunch and passage for passage of the present invention;
Fig. 3 is a passage bunch and the corresponding relation schematic diagram of time slot;
Fig. 4 is passage of the present invention bunch and time slot corresponding relation figure.
Embodiment
In the present invention, safeguard the passage that all passages of reflection take situation for each packaging time length in the mixed packaging time length and take table, when receiving the beginning Frame of a real time data stream when switching (take place passage user), the packaging time length that carries according to this frame is searched corresponding passage and is taken table, distributes passage by the situation that takies of passage in this table.A passage the simplest takies the corresponding relation of showing to comprise between the pairing channel number of passage and this passage seizure condition; Certainly also can increase other information as required.
Describe for convenience, introduce " passage bunch " and " time slot " two notions in the present invention.The one or more passages of passage bunch correspondence, the passage that removes maximum packaging time length correspondence takies off-balancesheet, and passage bunch takies port number corresponding in the table at different passages and is taken multiple (2 between pairing packaging time length of table and the maximum packaging time length by this passage
n, n=0,1,2...) determine.Time slot then is the time interval of data dispatching frame, the corresponding time slot of each passage in passage bunch.Time slot is minimum packaging time length in the present invention, therefore, concerns by the haplotype data between minimum packaging time length and the maximum packaging time length and to divide time slot.For example, mixed packaging time length comprises 20ms, 10ms and three packaging time lengths of 5ms, and the proportionate relationship of maximum packaging time length 20ms and minimum packaging time length 5ms is 4: 1, i.e. 4 times of relations.So:
A, the time interval are that the time slot of 5ms is divided into time slot 0, time slot 1, time slot 2 and time slot 3 totally 4 time slots, just use 4 time slots with all channel pollings one time;
B, take in the table at the passage of 5ms packaging time length correspondence, 4 passages of a passage bunch correspondence, promptly passage bunch takies 4 time slots;
C, take in the table at the passage of 10ms packaging time length correspondence, 2 passages of a passage bunch correspondence, promptly a passage family takies 2 time slots;
D, take in the table at the passage of 20ms packaging time length correspondence, a passage takies a time slot.
Except the passage of safeguarding packaging time length takies off-balancesheet, also to safeguard total passage bunch state table, be used to write down packaging time length, store the memory channel place passage bunch of this packaging time length real time data stream, and the corresponding relation between the shared time slot of memory channel.
When receiving the beginning Frame of a packing duration, the passage of searching this packaging time length correspondence takies table, calculate and be the distribution of flows memory channel of current packaging time length, and take in the table at each passage, with all path markings of this memory channel place passage bunch for taking, and in passage bunch state table the shared time slot of this memory channel of record.
After being packaging time length distribution of flows memory channel, calculating the Frame that receives according to packaging time length and current time slots and should write the position of memory channel, and be written in this memory channel.Simultaneously, in each time slot, according to the corresponding relation order poll passage of described passage bunch state table record, and from memory channel the sense data frame.
Be that example is elaborated with the mixed packaging time length that comprises 20ms, 10ms and three packaging time lengths of 5ms below:
Adopt the shared buffer memory of the SDRAM of 16M * 64 as voice packet, the address assignment of SDRAM as shown in Table 1:
Table one
| Address realm (24 hexadecimal addresses) | Function is distributed |
| 000000～0001FF | The voice packet of buffer memory passage 0 |
| 000200～0003FF | The voice packet of |
| 000400～0005FF | The voice packet of buffer memory passage 2 |
| 000600～0007FF | The voice packet of buffer memory passage 3 |
| ...... | ...... |
| FFFC00～FFFDFF | The voice packet of buffer memory passage 32K-2 |
| FFFE00～FFFFFF | The voice packet of buffer memory passage 32K-1 |
Wherein, sdram size is equally divided into 32K " OK " by 32K passage, and the 0th row SDRAM distributes to memory channel 0, the 1 row SDRAM and distributes to memory channel 1, the 2nd row SDRAM distributes to memory channel 2......, and the capable SDRAM of 32K-1 distributes to memory channel 32K-1.The capacity of each row SDRAM is 16M * 64/32K=512 * 64.
Adopt the SSRAM memory channel information of 1M * 32, the SSRAM capacity is equally divided into 32K " OK " by 32K passage, and the allocation tables of channel information is as shown in Table 2 among the SSRAM:
Table two
| Address realm (20 hexadecimal addresses) | Function is distributed |
| 00000～0001D | The channel information of memory channel 0 |
| 0001E | The mapping table of passage 0 |
| 0001F | The mapping table of memory channel 0 |
| 00020～0003D | The channel information of |
| 0003E | The mapping table of |
| 0003F | The mapping table of |
| ...... | ...... |
| FFFC0～FFFDD | The channel information of memory channel 32K-2 |
| FFFDE | The mapping table of passage 32K-2 |
| FFFDF | The mapping table of memory channel 32K-2 |
| FFFE0～FFFFD | The channel information of memory channel 32K-1 |
| FFFFE | The mapping table of passage 32K-1 |
| FFFFF | The mapping table of memory channel 32K-1 |
The memory channel of channel map recording channel correspondence and passage effective information in the table two, data format be as shown in Table 3:
Table three
| The position | Implication | Value |
| 31～20 | Keep and use | 0 |
| 19～16 | Passage packaging time length information | 0～15 |
| 15 | The passage | 1/0 |
| 14～0 | Memory channel number | 0～32767 |
Wherein, the 14th～0 of channel map position that in value is 0～32767 scope, is used to write down memory channel, the 15th effective information that is used for recording channel, value position 1 is that this passage is effective, value is that 0 o'clock this passage is invalid, the 18th～16 for being used to write down the packaging time length information of this passage, and the 31st～20 value is 0 to be used for keeping and to use.
The passage and the memory channel effective information of memory channel mapping table record memory channel correspondence in the table two, data format be as shown in Table 4:
Table four
| The position | Implication | Value |
| 31～16 | Keep and use | 0 |
| 15 | The memory channel | 1/0 |
| 14～0 | Passage | 0～32767 |
Wherein, the 14th～0 of memory channel mapping table position that in value is 0～32767 scope, is used to write down this memory channel, the 15th effective information that is used to write down this passage, value position 1 is that this passage is effective, value is that 0 o'clock this passage is invalid, and the 31st～16 value is 0 to be used for keeping and to use.
As previously mentioned, comprise that the proportionate relationship of maximum packaging time length 20ms and minimum packaging time length 5ms is 4: 1 in the mixed packaging time length of 20ms, 10ms and three packaging time lengths of 5ms, i.e. 4 times of relations.Therefore, under the 20ms packaging time length, maximum assignable port number (being port number) is 32k; Under the 10ms packaging time length, maximum assignable largest passages number of clusters is 16K; Under the 5ms packaging time length, maximum assignable largest passages number of clusters is 8K, and each passage takies in the table passage bunch and passage corresponding relation as shown in Figure 2.
The time interval is that the time slot of 5ms is divided into time slot 0, time slot 1, time slot 2 and time slot 3 totally 4 time slots, each time slot and passage bunch corresponding relation as shown in Figure 3, passage in 5ms packaging time length correspondence takies in the table, 4 time slots of each passage bunch correspondence, and take in the table 2 time slots of each passage bunch correspondence at the passage of 10ms packaging time length correspondence.In four time slots, each time slot poll finishes 8K passage.
Passage takies shown in the table five, and recording channel takies situation in the table.According to the difference of each packaging time length, its list item number difference.1 this passage of expression is bunch occupied, can not distribute use, and 0 this passage of expression bunch can distribute use.
Table five
| Passage bunch number | Taking table knows |
| Passage bunch 0 | 0 |
| | 1 |
| Passage bunches 2 | 1 |
| ...... | |
| Passage bunch 8K-1 | 0 |
Shown in the passage bunch state table six, record packaging time length in the table, distribute to the passage bunch number of the passage bunch correspondence of this packaging time length, and the corresponding relation between the pairing timeslot number of memory channel in this passage bunch (largest passages in the channel status table bunch calculate, be 8K-1 in this example) according to minimum packaging time length 5ms.
Table six
| Passage bunch number | Packaging time length | Timeslot number |
| Passage bunch 0 | 5ms | 00 |
| | 10ms | 00、01 |
| Passage bunches 2 | 20ms | 04 |
| ...... | ||
| Passage bunch 8K-1 |
When receiving the beginning Frame of packaging time length, distribute the process of passage and write data frame as follows:
If 1 receives the start frame (head bag) of 5ms, search the 5ms passage according to algorithm and take table, obtain the memory channel number of these voice, the passage of this memory channel place passage bunch correspondence taken all to be designated in the table knowledge table at 5ms, 10ms and 20ms passage take, and in passage bunch state table record; The calculated data frame should write the position of memory channel then, and is written in this memory channel.
If 2 receive the head bag of 10ms, search the 10ms passage according to algorithm and take table, obtain the memory channel number of these voice, take in the table at the 10ms passage this memory channel table is known for taking, take in the table at the 5ms passage passage bunch table of this memory channel correspondence is known for taking, take in the table gap marker with this memory channel place passage bunch correspondence for taking at the 20ms passage; The calculated data frame should write the position of memory channel then, and is written in this memory channel.
If 3 receive the head bag of 20ms, search the 20ms passage according to algorithm and take table, obtain the memory channel number of these voice, take in the table at the 20ms passage and this channel table known for taking, take in the table passage of this passage correspondence bunch is designated at 5ms, 10ms passage and take; The calculated data frame should write the position of memory channel then, and is written in this memory channel.
In order to simplify processing, if a passage A in the passage bunch X distributes to the voice flow that packaging time length is 5ms, this passage bunch X takies its excess-three passage corresponding in the table at the passage of 5ms packaging time length and can not reallocate to the voice flow of all packaging time lengths so, and soon four passages of passage bunch X correspondence take all to be labeled as in the table at all passages and take.If a channel B in the passage bunch Y is distributed to the 10ms voice flow, this passage bunch B takies another corresponding in table passage at the passage of 10ms packaging time length and can not reallocate to the voice flow of all packaging time lengths so, and two passages that are about to passage bunch Y correspondence take all to be labeled as in the table at all passages and take; Simultaneously, the passage of 5ms packaging time length take in the table with channel B in same passage bunch and the passage of 10ms packaging time length take in the table with channel B not two passages in same passage bunch can not reallocate to other packaging time lengths, but can distribute to the voice flow of 10ms packaging time length.If certain channel C is distributed to the 20ms voice flow, take in the table with channel C at the passage of 5ms packaging time length so and can not reallocate to the voice flow of other packaging time lengths, but can distribute to the voice flow of 20ms packaging time length at other three passages in the same passage bunch.Therefore, when reading, channel number can be defined as in " timeslot number+passage bunch number ", search thereby simplify.
After distributing passage and writing Frame, passage bunch and time slot corresponding relation are as shown in Figure 4.This Figure illustrates the writing position situation of 5ms, 10ms, 3 kinds of packaging time lengths of 20ms, when reading from passage bunch 0, time slot 0 beginning, level reads 8K passage bunch in proper order, changes time slot 1 then over to, again from passage bunch 0, until 8K-1 passage bunch, go on successively,, begin the next round circulation then up to time slot 3, passage bunch 8K-1.
According to channel number order poll, its process is as follows during read data frame:
Number search passage bunch state table according to the passage of current channel number correspondence bunch, obtain packaging time length, and carry out following operation according to different packaging time lengths:
If 1 this packaging time length is the 5ms packaging time length, then from passage bunch state table, read memory channel time slot corresponding number, and form channel number with " timeslot number+passage bunch number ", sense data frame from the memory channel of correspondence all will send 1 5ms voice packet of same 1 passage to output buffers to guarantee each time slot (being every 5ms time).
If 2 these packaging time lengths are the 10ms packaging time length, then from passage bunch state table, read memory channel time slot corresponding number, and form channel number with " timeslot number+passage bunch number ", sense data frame from the memory channel of correspondence sends 1 10ms voice packet of same 1 passage to output buffers to guarantee per 2 time slots (being every 10ms time).
If 3 these packaging time lengths are the 20ms packaging time length,, send 1 20ms voice packet of same 1 passage to output buffers to guarantee per 4 time slots (being every 20ms time) then according to the current channel number read data frame that is polled to.
Processing procedure behind the sense data frame is same as the prior art.
Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.Like this, if of the present invention these are revised and modification belongs within the scope of claim of the present invention and equivalent technologies thereof, then the present invention also is intended to comprise these changes and modification interior.
Claims (6)
1, a kind of implementation method of real-time data stream shaking reduction of mixed packaging time length; It is characterized in that this method comprises:
Safeguard step, be used to be respectively the passage that each packaging time length safeguards that reflection passage takies situation and take table;
Allocation step is used for when receiving the beginning Frame of a packing duration, searches corresponding passage and takies table and come distribution of flows memory channel for this packaging time length, and take in the table this memory channel is labeled as at all passages and take;
Write step is used to calculate the Frame that receives and should writes the position of memory channel, and is written in this memory channel;
Reading step is used for time interval of minimum packaging time length time slot as data dispatching order poll passage and sense data frame from memory channel.
2, the method for claim 1, it is characterized in that, concern according to the multiple between all the other each packaging time lengths and the maximum packaging time length except that maximum packaging time length, respective channel is taken passage in the table be divided into passage bunch, make the port number of each passage bunch correspondence equal described multiple, and each passage in each passage bunch take a time slot.
3, method as claimed in claim 2 is characterized in that, goes back maintenance channel bunch state table in safeguarding step, the corresponding relation between time slot that the memory channel of this state table record packaging time length, this packaging time length takies and the passage bunch number; Time slot that memory channel is shared, corresponding passage bunch number and corresponding packaging time length are recorded in this passage bunch state table in described allocation step.
4, as claim 2 or 3 described methods, it is characterized in that, in the described allocation step, to take at the passage of current packaging time length correspondence in the table to take all to be labeled as in the table at each passage in the rest channels in the same passage bunch and take, this rest channels can not be reallocated with the memory channel that distributes.
5, method as claimed in claim 4, it is characterized in that, in the described allocation step, when other packaging time length memory allocated passages of giving except that hand-to-hand fight in tights bag duration, to take at the passage of hand-to-hand fight in tights bag duration in the table with the memory channel that distributes in same passage bunch and take in the table and the described memory channel passage in same passage bunch not at other passages, take other passages the table at passage and take all to be labeled as in the table and take, make these passages no longer distribute to other packaging time lengths except that current packaging time length correspondence.
6, method as claimed in claim 3 is characterized in that, reading step specifically comprises:
Number inquire about described passage bunch state table according to the passage of the current channel number correspondence of poll bunch and obtain corresponding packaging time length;
Judge whether current packaging time length is maximum packaging time length, if, the pairing passage sense data of current channel number frame then from being polled to, otherwise, from passage bunch state table, obtain time slot corresponding and number form channel number, and from the pairing passage sense data of the channel number of this composition frame.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004101009048A CN100341267C (en) | 2004-12-02 | 2004-12-02 | Method for realizing real-time data stream shaking reduction of mixed packaging time length |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004101009048A CN100341267C (en) | 2004-12-02 | 2004-12-02 | Method for realizing real-time data stream shaking reduction of mixed packaging time length |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1783770A CN1783770A (en) | 2006-06-07 |
| CN100341267C true CN100341267C (en) | 2007-10-03 |
Family
ID=36773575
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2004101009048A Expired - Fee Related CN100341267C (en) | 2004-12-02 | 2004-12-02 | Method for realizing real-time data stream shaking reduction of mixed packaging time length |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100341267C (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105306958B (en) * | 2014-06-30 | 2018-08-17 | 惠州市伟乐科技股份有限公司 | A kind of method and device of IP-based transmission stream bit rate smoothing processing |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002271388A (en) * | 2001-03-12 | 2002-09-20 | Iwatsu Electric Co Ltd | Delay buffer control system for real time packets |
| CN1453949A (en) * | 2002-04-26 | 2003-11-05 | 东方通信科技发展有限公司 | Method of eliminating business flow jitter and slip code in CDMA communication system |
| CN1463125A (en) * | 2002-05-28 | 2003-12-24 | 华为技术有限公司 | Large capacity realtime stream processing method for removing dithering in using buffer memory |
| US20040120309A1 (en) * | 2001-04-24 | 2004-06-24 | Antti Kurittu | Methods for changing the size of a jitter buffer and for time alignment, communications system, receiving end, and transcoder |
-
2004
- 2004-12-02 CN CNB2004101009048A patent/CN100341267C/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002271388A (en) * | 2001-03-12 | 2002-09-20 | Iwatsu Electric Co Ltd | Delay buffer control system for real time packets |
| US20040120309A1 (en) * | 2001-04-24 | 2004-06-24 | Antti Kurittu | Methods for changing the size of a jitter buffer and for time alignment, communications system, receiving end, and transcoder |
| CN1453949A (en) * | 2002-04-26 | 2003-11-05 | 东方通信科技发展有限公司 | Method of eliminating business flow jitter and slip code in CDMA communication system |
| CN1463125A (en) * | 2002-05-28 | 2003-12-24 | 华为技术有限公司 | Large capacity realtime stream processing method for removing dithering in using buffer memory |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1783770A (en) | 2006-06-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5640234B2 (en) | Layer 2 packet aggregation and fragmentation in managed networks | |
| US7730238B1 (en) | Buffer management method and system with two thresholds | |
| CN1275433C (en) | Method and equipment for transfering variable synchronous data | |
| US7653072B2 (en) | Overcoming access latency inefficiency in memories for packet switched networks | |
| CN111327391B (en) | Time division multiplexing method, device, system and storage medium | |
| US6496899B1 (en) | Disk scheduling system with bounded request reordering | |
| US20020169921A1 (en) | Packet buffer | |
| JPH08510071A (en) | Queue management device and method | |
| CN1109231A (en) | Method and apparatus for temporarily storing data packet and switch using the same | |
| US7587549B1 (en) | Buffer management method and system with access grant based on queue score | |
| RU2009148525A (en) | DIGITAL DATA MANAGEMENT USING A MEMORY USED POOL | |
| CN1689283A (en) | Method and apparatus for processing superframes using an arbitration system | |
| US7865632B2 (en) | Memory allocation and access method and device using the same | |
| CN1529962A (en) | System and method for processing bandwidth allocation messages | |
| JP2005500620A (en) | Memory pool with moving memory blocks | |
| US7760736B2 (en) | Method, system, and computer program product for ethernet virtualization using an elastic FIFO memory to facilitate flow of broadcast traffic to virtual hosts | |
| US7565496B2 (en) | Sharing memory among multiple information channels | |
| CN114500401B (en) | Resource scheduling method and system for coping with burst traffic | |
| CN100341267C (en) | Method for realizing real-time data stream shaking reduction of mixed packaging time length | |
| CN1214581C (en) | Large capacity realtime stream processing method for removing dithering in using buffer memory | |
| KR20130098265A (en) | Computer system and method for operating the same | |
| CN1601501A (en) | Memory bandwidth control device | |
| CN1288876C (en) | Dynamic RAM quene regulating method based on dynamic packet transmsision | |
| US7751400B2 (en) | Method, system, and computer program product for ethernet virtualization using an elastic FIFO memory to facilitate flow of unknown traffic to virtual hosts | |
| US20080181247A1 (en) | Method and apparatus for processing queue in network system |
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: 20071003 Termination date: 20121202 |