CN105812097B - An adaptive adjustment method of AMR coding rate based on network state - Google Patents
An adaptive adjustment method of AMR coding rate based on network state Download PDFInfo
- Publication number
- CN105812097B CN105812097B CN201610148403.XA CN201610148403A CN105812097B CN 105812097 B CN105812097 B CN 105812097B CN 201610148403 A CN201610148403 A CN 201610148403A CN 105812097 B CN105812097 B CN 105812097B
- Authority
- CN
- China
- Prior art keywords
- cmr
- value
- delay
- current
- sender
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 23
- 230000003044 adaptive effect Effects 0.000 title description 10
- 230000033228 biological regulation Effects 0.000 claims abstract description 9
- 230000005540 biological transmission Effects 0.000 claims description 6
- 230000008859 change Effects 0.000 abstract description 8
- 230000006854 communication Effects 0.000 description 15
- 238000004891 communication Methods 0.000 description 12
- 238000012360 testing method Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 230000006978 adaptation Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 4
- 238000010295 mobile communication Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000006855 networking Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000035772 mutation Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000011056 performance test Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000001303 quality assessment method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 230000035807 sensation Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/0001—Systems modifying transmission characteristics according to link quality, e.g. power backoff
- H04L1/0014—Systems modifying transmission characteristics according to link quality, e.g. power backoff by adapting the source coding
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Telephonic Communication Services (AREA)
Abstract
A kind of AMR code rate self-adapting regulation method based on network state of the present invention, it is specific as follows: one, to determine initial code mode for the sender of call；Two, RTP packet is written into the corresponding rate of the initial code mode of sender；Three, RTP packet is transferred to receiving end, and receiving end calculates the packet loss and delay parameter of sender；Four, receiving end utilizes packet loss and delay parameter, judges the current initial code rate of the sender network state current if appropriate for receiving end, if be suitble to, keeps current code rate；Otherwise, change CMR value according to speed adjust strategy；Five, by the corresponding rate write-in RTP packet output of CMR value adjusted to sender, sender similarly calculates packet loss and delay parameter, and is judged by adjustable strategies, changes CMR value, and then adjust AMR code rate in real time for receiving end；Advantage is: by adjusting code rate in real time, can adapt to the continuous variation of Network status, to guarantee the good call experience of user.
Description
Technical field
The invention belongs to mobile Internet speech processes fields, describe a kind of AMR code rate based on network state
Self-adapting regulation method.
Background technique
Continuous development with mobile communication technology and Internet technology with merge, mobile information and people's lives produce
Increasingly closer connection is given birth to, it is desirable to easily access internet obtains information and service whenever and wherever possible.
In by the end of August, 2015 by, the mobile interchange network users at existing 9.46 hundred million families of China, the sum of surfing Internet with cell phone is more than 9
Hundred million, the ratio for accounting for mobile phone user is up to 69.5%.Largely the application based on mobile Internet and business are extended to from amusement
Every aspect in social life.With the development of network technology and the sustainable growth of mobile data flow, voice calling service
The limitation for breaking through conventional telecommunications telephone service, the networking telephone (the Voice over high towards call tone quality, rate are low
Internet Protocol, VoIP) direction develop.But in the case where Network status is unstable, the net based on mobile terminal
There are still problem, voice call qualities to be guaranteed well for network phone.
Self-adapting multi-rate narrowband voice coding (Adaptive Multi Rate-Narrow Band Speech
Codec, AMR-NB) algorithm be based on Code Excited Linear Prediction (Code Excited Linear Prediction, CELP) calculate
The speech coder standard of method is mainly used in 3G (Third Generation) Moblie wideband code division multiple access (Wideband Code
Division Multiple Access, W-CDMA) in system.
A variety of coding modes that AMR coding provides can better solve the rate allocation of source and channel coding,
Certain coding mode can be adaptive selected according to channel conditions to be transmitted, so that the distribution and utilization to radio resource are more
Adduction is managed, effectively, flexibly, meanwhile, preferably call can be also provided for user and experienced.
The frame structure of AMR is as shown in Figure 1, include 3 parts: AMR frame head, AMR auxiliary information and AMR core frame data；Its
In, AMR frame head includes the frame type of 4bits and the frame feature indicating bit of 1bit.AMR auxiliary information is used for mode adaptive and mistake
Detection, specifically includes: the mode flags of 3bits, the mode request of 3bits and the coding cycle redundancy check of 8bits.AMR core
Heart frame data include voice or noise data, wherein the data of each frame are divided into: A/B/C three classes field, A class field are in a frame
Most important data, once A class field is damaged, entire frame can not just be decoded.So generally to make when wireless transmission
A class field portions data are protected with the method for various redundancy checks.Compared with A class field, B class field be relatively not that
Important data；C class field is to influence the smallest data to whole decoding.
AMR speech coder can effectively improve voice quality, enhance the ability of the anti-channel errors of system, while increasing and being
System capacity, has the advantages that other voice coding modes are irreplaceable.Meanwhile AMR can be selected different according to Network status
Coding mode can better adapt to the variation of network.
2015, the country had more than hundred sections of network telephone software, as people are to the degree of dependence of smart phone
Increasing, the networking telephone also develops towards cell phone end, currently more popular network telephone software, such as: it is wechat phone directory, micro-
Words, 360 freephones, cloud such as exhale at the version for being also all proposed mobile terminal, to adapt to the communication requirement of people.Although most of
The speech quality of the mobile terminal networking telephone substantially and black phone is suitable, but in the case where Network status is unstable, is still deposited
In the unstable problem of speech quality, call experience well cannot be provided for people.
Summary of the invention
The present invention analyzes and researches regarding to the issue above, and the audio communication system based on mobile Internet is compiled using AMR
Code mode, different code rates is selected according to Network status, and preferably solving the unstable bring of Network status influences, and is
This proposes a kind of AMR code rate self-adapting regulation method based on network state, the specific steps are as follows:
Step 1: for the sender of call, the current network type according to locating for oneself determines initial code mode；
Current network type locating for sender includes: 2G, 3G, 3G or more or WiFi；
Code mode request CMR (Codec Mode Request), shares 8 values, is followed successively by 0,1,2,3,4,5,6 and 7；
If current sender connection is that 2G type network is conversed, initial code mode is CMR=0；
If the network type of current sender connection is 3G, 3G or more or WiFi, then initial code mode is CMR=
7；
Step 2: RTP packet is written in the corresponding rate of the initial code mode of sender；
The corresponding rate of initial code mode is successively are as follows: the lower rate of mode 0 is 4.75kbit/s；Rate is under mode 1
5.15kbit/s；The lower rate of mode 2 is 5.90kbit/s；The lower rate of mode 3 is 6.70kbit/s；Rate is under mode 4
7.40kbit/s；The lower rate of mode 5 is 7.95kbit/s；The lower rate of mode 6 is 10.2kbit/s；Rate is under mode 7
12.2kbit/s。
Initial CMR is there are two types of state, and AMR code rate is arranged in 4.75kbps when CMR=0；AMR coding speed when CMR=7
Rate is arranged in 12.2kbps.
Step 3: RTP packet is transferred to the receiving end of call, receiving end calculates the packet loss and delay parameter of sender；
The packet loss cal_lost of RTP packet calculates as follows:
Cal_lost=(packetlost/good*100)
Packetlost is packet loss quantity；Good is the RTP packet quantity of transmission.
Add up instantaneous time delay value cal_delay:
Cal_delay+=(jitter-250*mu)/8/mu；
Jitter is delay variation；Mu is sample rate, is the multiple of 8K；
It is primary that judgement in every 1 second or so is arranged in average delay avg_delay:
Avg_delay=cal_delay/timeCount
TimeCount is frequency of reckoning by time, and every 1 second is 1 time.
According to delay variation statistical value, final time delay D is calculatedtr:
Dtr(n+1)=α Dtr(n)+(1-α)cal_delay
Wherein Dtr(n) be last Delay Variation value.
Step 4: receiving end utilizes packet loss and delay parameter, judge whether the current initial code rate of sender fits
The current network state of splice grafting receiving end keeps current code rate if be suitble to；Otherwise, changed according to speed adjust strategy
Become CMR value；
Specific step is as follows:
Step 401, according to time delay level, two delay thresholds: Dth1 and Dth2 are set；
Step 402 judges whether time delay meets: Dtr< Dth1, if so, entering step 403；Otherwise 407 are entered step；
Step 403 judges whether packet loss meets cal_lost≤1%, if so, entering step 404；Otherwise, enter
Step 405；
Step 404 judges whether CMR value is 7；If so, keeping CMR value constant；Otherwise, current CMR value increases 1 certainly；
Step 405 further judges whether packet loss meets: 1% < cal_lost≤5%, if so, entering step
406, otherwise current CMR value is set as 0；
Step 406 judges whether CMR value is more than or equal to 1；If so, current CMR value subtracts 1 certainly；Otherwise, by current CMR value
It is set as 0；
Step 407 further judges whether time delay meets Dth1≤Dtr≤ Dth2, if so, entering step 408；Otherwise,
Current CMR value is set as 0；
Step 408 judges whether packet loss meets cal_lost≤1%, if so, entering step 409；Otherwise, enter
Step 410；
Step 409 judges whether CMR value is more than or equal to 1；If so, current CMR value subtracts 1 certainly；Otherwise, by current CMR value
It is set as 0；
Step 410 judges whether packet loss meets 1% < cal_lost≤5%, if so, entering step 411；Otherwise
Current CMR value is set as 0；
Step 411 judges whether CMR value is more than or equal to 2；If so, current CMR value subtracts 2 certainly；Otherwise, by current CMR value
It is set as 0；
Step 5: receiving end exports the corresponding rate write-in RTP packet of CMR value adjusted to sender, sender is same
Reason calculates packet loss and delay parameter, and is judged by the adjustable strategies of step 4, changes CMR value, and then adjust AMR in real time
Code rate.
The present invention has the advantages that
1, a kind of AMR code rate self-adapting regulation method based on network state makes full use of calculation by given threshold
Method parameter is few and realizes simple characteristic, fine can obtain and react to the variation of Network status.
2, a kind of AMR code rate self-adapting regulation method based on network state, by adjusting code rate in real time, from
And guarantee the good call experience of user.
Detailed description of the invention
Fig. 1 is the frame structure schematic diagram of AMR of the present invention；
Fig. 2 is AMR speech coding mode schematic diagram of the present invention；
Fig. 3 is that GSM of the present invention works in the relational graph under full-rate mode between MOS value and C/I value；
Fig. 4 is that GSM of the present invention works in the relational graph under half-rate mode between MOS value and C/I value；
Fig. 5 is that AMR adaptive coding code rate of the present invention switches schematic diagram；
Fig. 6 is a kind of AMR code rate self-adapting regulation method flow chart based on network state of the present invention；
Fig. 7 is RTP packet format of the present invention and Payload Header form schematic diagram；
Fig. 8 is the method flow diagram that the present invention judges initial code rate if appropriate for current network state；
Fig. 9 is the interface display diagram in communication process of the present invention.
Specific embodiment
Below in conjunction with attached drawing, the present invention is described in further detail.
In the existing mobile communication system of AMR, the principle of adaptation mechanism is substantially the same.With global system for mobile communications
It is analyzed for GSM (Global System for Mobile Communication)；
Using the AMR speech coding mode used in GSM as shown in Fig. 2, the coding mode used is divided into full rate
(Full Rate, FR) and two kinds of situations of half rate (Half Rate, HR).GSM is when selecting coding mode first according to load
The dry value than (Interference/Carrier, C/I) selects coding mode；In the preferable situation of carrier/interface ratio, the volume of selection
Pattern is AMR-HR, can provide higher wireless capacity.And in the case where C/I is poor, then AMR-FR is worked in, is provided
Sound quality in better words than enhanced full rate；
Under the conditions of full-rate mode and under the conditions of half-rate mode, each code rate is surveyed under conditions of different C/I values
Fixed MOS Distribution value figure is as shown in Figure 3 and Figure 4, it can be seen that with the reduction of carrier interference ratio C/I value, the MOS value of each code rate is all
Reduced trend is shown as, because with the enhancing of interference and noise, the communication effect and speech quality that human ear is heard will become
Difference.
It is as shown in Figure 5 that AMR adaptive coding code rate switches schematic diagram, it is known that in gsm system AMR adaptation mechanism,
The buffer area being arranged when switching between adjacent encoder rate, with the variation of C/I value, code rate will not be immediately switched to
Adjacent code rate, but the switching of code rate is just carried out after the threshold value beyond buffer area.Using buffer area, speed can be prevented
The frequent jump of rate.
A kind of AMR code rate self-adapting regulation method based on network state is adopted in open source SIP software Sipdroid
Using the AMR code rate adaptive re-configuration police based on threshold method is and to introduce time delay using time delay and packet loss as parameter and become
The considerations of gesture is predicted, has formulated rate adaptation adjustable strategies, the real-time change of Lai Shiying network.
Overall flow figure is as shown in Figure 6, the specific steps are as follows:
Step 1: for the sender of call, the current network type according to locating for oneself determines initial code mode；
Current network type locating for each sender includes: 2G, 3G, 3G or more or WiFi；
Code mode request CMR (Codec Mode Request), shares 8 values, is followed successively by 0,1,2,3,4,5,6 and 7；
Using algorithm of the monitoring without the channel RTP, the initial volume of each sender is determined using the network type for receiving terminal
Bit rate.If current sender connection is that 2G type network is conversed, initial code mode is CMR=0；Initial speed
Rate will be arranged in lower grade.If the time delay of recipient, packet loss are all smaller, can code rate to sender it is upward
Adjustment., whereas if the network type of current sender connection is 3G, 3G or more or WiFi network, then initial code mode
For CMR=7；Initial code rate is arranged in higher grade, if the time delay of recipient, packet loss are bigger, to downward
Reorganize bit rate.Wherein adjustment downwardly and upwardly is reversible, and is determined the case where with specific reference to time delay and packet loss.
Step 2: RTP packet is written in the corresponding rate of the initial code mode of sender；
The corresponding rate of initial code mode is successively are as follows: the lower rate of mode 0 is 4.75kbit/s；Rate is under mode 1
5.15kbit/s；The lower rate of mode 2 is 5.90kbit/s；The lower rate of mode 3 is 6.70kbit/s；Rate is under mode 4
7.40kbit/s；The lower rate of mode 5 is 7.95kbit/s；The lower rate of mode 6 is 10.2kbit/s；Rate is under mode 7
12.2kbit/s。
Specific transmission process is as follows: terminal 1 is conversed with terminal 2, and terminal 1 is used as sender, sends initial code speed
For rate to terminal 2, initial Mode is passed to the encode function of terminal 1, and FT field in RTP packet is written in Mode；Terminal 1 writes CMR value
It is sent in the first character section of payload, terminal 2 is used as receiving end, and decode function will be in the RTP packet that received before
The first character section of payload takes out, and obtains CMR value, obtains Mode according to CMR；Terminal 2 counts the time delay that terminal 1 sends packet
And packet loss, 1 present rate mode of terminal is determined if appropriate for the network of terminal 2, if improper, by end after adjustment CMR
End 2 is sent, and the payload header in RTP packet notifies terminal 1, at this point, terminal 2 is used as sender, similarly, terminal 1 is received
To terminal 2 send RTP packet unpacked, after according to CMR value progress code rate adjustment.
According to the demand of transmission plan, call terminal is needed to judge the network type at place, and provides initial CMR
Value.
Network type of the sending side terminal according to locating for oneself carries out the determination of initial CMR, state that there are two types of initial CMR,
The CMR=0 under 2G network, i.e. AMR code rate are arranged in 4.75kbps, in 3G, CMR under the conditions of 3G or more or WiFi network
=7, i.e. AMR code rate is arranged in 12.2kbps.
Step 3: RTP packet is transferred to receiving end, receiving end calculates the packet loss and delay parameter of sender.
Provide that RTP packet format is as shown in fig. 7, simultaneously about the RTP transformat of AMR according in RFC3267 and RFC4867
It notifies receiving end, receiving end to unpack after receiving RTP packet by the packet header of RTP packet, extracts CMR value,
Wherein RTP header has 12 bytes, and Payload Header has a byte, using octet-align
Mode, Table of Content contain the head of AMR frame, and subsequent part is effective speech frame.And in Sipdroid
12 bytes of middle RTP Header are not include coded modulation in the payload of this part followed by payload payload
Rate (Codec Mode Request, CMR) field.Therefore it needs to do certain place to the RTP packet format in Sipdroid
Reason: in originator plus a byte, byte that receiving end correspondence subtracts one thus may include the judgement information of code rate.By Payload
Header is designed as octet-align mode, i.e. Gao Siwei represents CMR value, and low four are left 0.
The time delay given out a contract for a project according to receiving end to transmitting terminal, packet drop, devise threshold method adjustable strategies.
Packet loss is the principal element considered, and time delay is another Consideration.It is root for the first time when adjustment code rate
Judged according to initial CMR, later during, the judgement of CMR is adjusted according to the previous CMR value for determining the moment
's.
Firstly, being utilized respectively the packet loss cal_lost that code statement calculates RTP VoP:
Cal_lost=(packetlost/good*100)
Packetlost is packet loss quantity；Good is the data packet number of transmission.
Accumulated delay value cal_delay:
Cal_delay+=(jitter-250*mu)/8/mu；
Jitter is delay variation；Mu is sample rate, is the multiple of 8K；If sample rate is 16k, mu 2；
It is primary that judgement in every 1 second or so is arranged in average delay avg_delay:
Avg_delay=cal_delay/timeCount
TimeCount is frequency of reckoning by time, and every 1 second is 1 time.
Since Network status can have the case where mutation, time delay obtained above may be an instantaneous value.In network
In tolerable situation, the value of packet loss and time delay may change rapidly, therefore according to average packet loss ratio and instantaneously
Time delay value is possible inaccurate to carry out code rate selection.Such as certain flashy packet loss and time delay it is lower, at this time just at
The time of judgement, so that the code rate of 12.2k has been selected, and actually network environment is bad, may cause packet loss and time delay
Increase, deteriorates communication effect.
It, thus can be to avoid the variation of instantaneous time delay therefore, it is necessary to judge the variation tendency of time delay whithin a period of time
Influence the selection of code rate.According to the calculation formula about Delay Variation trend proposed in document, and pass through actual test,
The considerations of having chosen reasonable weighted value, joined Delay Variation trend in rate adjustment algorithm, is calculated using calculation formula
Time delay.It thus can effectively solve the problems, such as to jump frequent, and make the adjustment algorithm of rate more reasonable.
The case where instantaneously jumping in view of time delay, this paper presents the processing methods for being directed to delay variation, when adding one
Prolong shake statistical value N, when current time delay and previous time delay difference DELTA d are more than three times continuously threshold delta T=80ms, i.e. N=3,
Prove that Network status is persistently deteriorated at this time, it is therefore desirable to more consider the time delay value obtained at present, obtain as follows
Time-delay calculation formula:
Dtr(n+1)=α Dtr(n)+(1-α)cal_delay
Wherein Dtr(n) be last Delay Variation value, Dtr(n+1) it is value next time.The meaning of this formula is just
It is to eliminate instantaneous i.e. influence of the amplitude of variation very greatly to code rate judgement of time delay value shake, DtrAll with pervious time delay value
It is related, there is Memorability, be the value of a consecutive variations.
According to actual measurement, the considerations of adding to delay variation statistical value, as N=3, i.e., when Network status is persistently deteriorated, α
=0.7, the weight of instantaneous time delay is higher.As N < 3, it was demonstrated that time delay at this time is mutated only temporary, instantaneous time delay weight
Smaller α=0.9.Above-mentioned value is according to the empirical value that repeatedly test is chosen.
Step 4: receiving end judges the current initial code of sender using packet loss and delay parameter according to RTP packet
Rate keeps current code rate if be suitble to if appropriate for the current network state of receiving end；Otherwise, it is real to change CMR value
When adjust AMR code rate.
The every 1s in receiving end carries out primary calculate and judges；
As shown in Figure 8, the specific steps are as follows:
Step 401, according to time delay level, two delay thresholds: Dth1 and Dth2 are set；
Dth1 is 200ms, Dth2 600ms；Due to time delay be lower than 200ms when, the quality of voice communication hardly by
It influences, when time delay is more than 600ms, to the obvious effect of voice communication.And Dth1 is selected to set according to actual measurement experience
It is set to 200ms, Dth2 is set as 600ms.
Step 402 judges whether time delay meets: Dtr< Dth1, if so, entering step 403；Otherwise 407 are entered step；
(it is lower than threshold value Dth1) when time delay is in reduced levels, AMR code rate is only determined by packet loss.
Step 403 judges whether packet loss meets cal_lost≤1%, if so, entering step 404；Otherwise, enter
Step 405；
The threshold value of two packet loss is horizontally disposed with according to packet loss, since when packet loss is lower than 1%, AMR voice is logical
Words quality is preferable, and both-end is conversed according to initial code rate, illustrates that Network status is good at this time, can be from initial CMR
Value improves step by step.
Step 404 judges whether CMR value is 7；If so, keeping CMR value constant；Otherwise, current CMR value increases 1 certainly；
Step 405 further judges whether packet loss meets: 1% < cal_lost≤5%, if so, entering step
406, otherwise current CMR value is set as 0；
When packet loss appears between 1%~5%, voice call quality will be affected.At this point, according to current
The case where CMR value, drops level-one to CMR accordingly.It is poor that Network status is turned out when packet loss is more than 5%, directly selects CMR=
0 minimum code rate 4.75kbps.
Step 406 judges whether CMR value is more than or equal to 1；If so, current CMR value subtracts 1 certainly；Otherwise, by current CMR value
It is set as 0；
Step 407 further judges whether time delay meets Dth1≤Dtr≤ Dth2, if so, entering step 408；Otherwise,
Current CMR value is set as 0；
When time delay level is between threshold value Dth1 and Dth2, it was demonstrated that network delay is larger at this time, has to speech quality
Certain influence illustrates that Network status is in poor state when time delay, which is in, is greater than delay threshold Dth2, therefore, it is determined that
CMR=0 is conversed using the minimum code rate 4.75kbps of AMR.
Step 408 judges whether packet loss meets cal_lost≤1%, if so, entering step 409；Otherwise, enter
Step 410；
When packet loss is lower than 1%, due to the influence of time delay, the rank of CMR can also be suitably turned down, is reduced step by step.
Step 409 judges whether CMR value is more than or equal to 1；If so, current CMR value subtracts 1 certainly；Otherwise, by current CMR value
It is set as 0；
Step 410 judges whether packet loss meets 1% < cal_lost≤5%, if so, entering step 411；Otherwise
Current CMR value is set as 0；
When packet loss appears between 1%~5%, voice communication is affected by packet loss, while time delay is also right
Speech quality impacts, and at this moment two-stage drops in CMR.
Step 411 judges whether CMR value is more than or equal to 2；If so, current CMR value subtracts 2 certainly；Otherwise, by current CMR value
It is set as 0；
By the processing to time delay and packet loss, so that Adaptive Rate Shape strategy efficiently avoids network mutation and brings
Decision error, advantageously ensure that AMR call effect.Along with the mode of the increasing or decreasing step by step of code rate, reduce
The shake of code rate also avoids occurring code rate improves suddenly excessive and generation the case where cause network congestion,
Improve the quality of voice communication.The code rate mode of present terminal exports on the screen, the interface display in communication process
As shown in Figure 9.
Step 5: receiving end exports the corresponding rate write-in RTP packet of CMR value adjusted to sender, sender is same
Reason calculates packet loss and delay parameter, and is judged by the adjustable strategies of step 4, changes CMR value, and then adjust AMR in real time
Code rate.
When carrying out the call of AMR codec speech, software can determine initial compile according to the current network conditions of acquisition first
Bit rate can be adaptively adjusted code rate according to the real-time change of Network status, with preferably then in communication process
Using existing network condition, preferably call experience is provided for user.
Voice communication test is carried out to threshold method rate adaptation adjustable strategies, to prove the superiority of its performance.This survey
Criterion is done in examination using average opinion value (Mean Opinion Score, MOS), and it is flat that MOS value is primarily referred to as voice quality
Equal index is exactly the voice template comparison in the voice and MOS instrument that test calls are collected.Score value is 1~5 point, main to reflect
Be user perceptibility, general and signal intensity, disturbed condition, switch instances are related, and 5 points are highest.Test MOS value is adopted
With subjective and objective two kinds of test methods, wherein subjective MOS using ITU-T P.800 and P.830 recommendation, by different people
There is the corpus of decline to carry out subjective sensation comparison to original language material and after system is handled respectively, obtains MOS points, finally ask flat
Mean value.
Objective MOS evaluation is then using the ITU-T subjective speech quality assessment (Perceptual that P.862 recommendation provides
Evaluation of Speech Quality, PESQ) method, it is tested by special software PESQ.PESQ and MOS value
Corresponding relationship is as shown in table 1.
Table 1
| Rank | MOS value | User satisfaction | PESQ |
| It is excellent | 5 | It is very good, listen it is clear that undistorted sense, non-delay sense | 3.5-4.5 |
| It is good | 4 | It is slightly worse, listen clear, delay is small, there is a noise | 3.0-3.5 |
| In | 3 | It is not clear have certain delay, there is noise, there is distortion it is also possible that listening | 2.4-3.0 |
| Difference | 2 | Reluctantly, it listens less clear, has compared with loud noise or discontinuously, distortion is serious | 1.7-2.4 |
| It is bad | 1 | Very poor, mute or can not hear clearly completely, noise is very big | 1.0-1.7 |
Test passes through two ways, subjectivity evaluation and test and objective examination.Subjectivity evaluation and test be using call software Sipdroid into
Row dual end communication, caller score to speech quality according to MOS value evaluating standard.Objective examination is to design machine in software
The voice recording of both call sides is got off to be stored in mobile phone by system, and then both sides' voice is scored by test and evaluation software PESQ.
Obtain the speech performance test result such as table 2.
Table 2
It can be seen from data in table 2 effect of the threshold method rate adaptation adjustable strategies when realizing voice communication compared with
Good, MOS value score value is higher, does not have significant difference with the MOS value of other rates.This proves threshold method rate adaptation adjustable strategies
Better performances, can be realized as the real-time change of network is in time adjusted speech encoding rate, and guarantee converse matter
Amount.
The invention proposes the AMR code rate adaptive re-configuration polices based on threshold method, realize based on mobile interchange
The audio communication system of net makees parameter using time delay and packet loss, and introduce time delay according to the real-time change situation of Network status
The considerations of trend prediction, dynamically adjusts ARM code rate, the real-time change of Lai Shiying network；So that AMR voice coding is conversed
According to Network status real-time perfoming Adaptive Rate Shape, to provide preferably call experience for user.
Claims (1)
1. a kind of AMR code rate self-adapting regulation method based on network state, which is characterized in that specific step is as follows:
Step 1: for the sender of call, the current network type according to locating for oneself determines initial code mode；
If current sender connection is that 2G type network is conversed, initial code mode is CMR=0；
If the network type of current sender connection is 3G, 3G or more or WiFi, then initial code mode is CMR=7；
Step 2: RTP packet is written in the corresponding rate of the initial code mode of sender；
AMR code rate is arranged in 4.75kbps when CMR=0；AMR code rate is arranged in 12.2kbps when CMR=7；
Step 3: RTP packet is transferred to the receiving end of call, receiving end calculates the packet loss and delay parameter of sender；
The packet loss cal_lost of RTP packet calculates as follows:
Cal_lost=(packetlost/good*100)
Packetlost is packet loss quantity；Good is the RTP packet quantity of transmission；
Add up instantaneous time delay value cal_delay:
Cal_delay+=(jitter-250*mu)/8/mu；
Jitter is delay variation；Mu is sample rate, is the multiple of 8K；
It is primary that judgement in every 1 second is arranged in average delay avg_delay:
Avg_delay=cal_delay/timeCount
TimeCount is frequency of reckoning by time, and every 1 second is 1 time；
According to delay variation statistical value, final time delay D is calculatedtr:
Dtr(n+1)=α Dtr(n)+(1-α)cal_delay
Wherein Dtr(n) be last Delay Variation value；
Step 4: receiving end utilizes packet loss and delay parameter, judge the current initial code rate of sender if appropriate for connecing
The current network state of receiving end keeps current code rate if be suitble to；Otherwise, CMR is changed according to speed adjust strategy
Value；Specifically:
Step 401, according to time delay level, two delay thresholds: Dth1 and Dth2 are set；
Step 402 judges whether time delay meets: Dtr< Dth1, if so, entering step 403；Otherwise 407 are entered step；
Step 403 judges whether packet loss meets cal_lost≤1%, if so, entering step 404；Otherwise, it enters step
405；
Step 404 judges whether CMR value is 7；If so, keeping CMR value constant；Otherwise, current CMR value increases 1 certainly；
Step 405 further judges whether packet loss meets: 1% < cal_lost≤5%, if so, 406 are entered step, it is no
Then current CMR value is set as 0；
Step 406 judges whether CMR value is more than or equal to 1；If so, current CMR value subtracts 1 certainly；Otherwise, current CMR value is set as
0；
Step 407 further judges whether time delay meets Dth1≤Dtr≤ Dth2, if so, entering step 408；Otherwise, will work as
Preceding CMR value is set as 0；
Step 408 judges whether packet loss meets cal_lost≤1%, if so, entering step 409；Otherwise, it enters step
410；
Step 409 judges whether CMR value is more than or equal to 1；If so, current CMR value subtracts 1 certainly；Otherwise, current CMR value is set as
0；
Step 410 judges whether packet loss meets 1% < cal_lost≤5%, if so, entering step 411；Otherwise current
CMR value is set as 0；
Step 411 judges whether CMR value is more than or equal to 2；If so, current CMR value subtracts 2 certainly；Otherwise, current CMR value is set as
0；
Step 5: receiving end exports the corresponding rate write-in RTP packet of CMR value adjusted to sender, sender is similarly counted
Packet loss and delay parameter are calculated, and carries out judgement by the adjustable strategies of step 4 and changes CMR value, and then adjust AMR coding in real time
Rate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610148403.XA CN105812097B (en) | 2016-03-16 | 2016-03-16 | An adaptive adjustment method of AMR coding rate based on network state |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610148403.XA CN105812097B (en) | 2016-03-16 | 2016-03-16 | An adaptive adjustment method of AMR coding rate based on network state |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105812097A CN105812097A (en) | 2016-07-27 |
| CN105812097B true CN105812097B (en) | 2019-04-23 |
Family
ID=56468490
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610148403.XA Active CN105812097B (en) | 2016-03-16 | 2016-03-16 | An adaptive adjustment method of AMR coding rate based on network state |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105812097B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107666366B (en) * | 2016-07-28 | 2020-02-14 | 华为技术有限公司 | Method, device and system for adjusting coding rate |
| CN109036444A (en) * | 2018-08-21 | 2018-12-18 | 京信通信系统（中国）有限公司 | Speech encoding rate method of adjustment, device, computer storage medium and equipment |
| CN112398702B (en) * | 2019-08-12 | 2022-08-02 | 腾讯科技（深圳）有限公司 | Speed measurement method and device, server and storage medium |
| CN112202618A (en) * | 2020-02-23 | 2021-01-08 | 徐世云 | Equipment state adjusting method based on data communication and data processing terminal |
| CN113784338B (en) * | 2020-05-21 | 2023-05-26 | 成都鼎桥通信技术有限公司 | Voice service bearing method in multi-service broadband cluster system |
| CN112565884B (en) * | 2020-11-27 | 2024-03-05 | 北京达佳互联信息技术有限公司 | Image processing method, device, terminal, server and storage medium |
| CN112866746A (en) * | 2020-12-31 | 2021-05-28 | 杭州雾联科技有限公司 | Multi-path streaming cloud game control method, device, equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436990A (en) * | 2008-12-23 | 2009-05-20 | 深圳华为通信技术有限公司 | Method for automatically adjusting encoding rate, receiving device and communication system |
| CN103888846A (en) * | 2014-03-04 | 2014-06-25 | 浙江大学 | Wireless video streaming service self-adaption rate control method based on QoE |
| CN104009819A (en) * | 2013-02-22 | 2014-08-27 | 中兴通讯股份有限公司 | Anti-packet-loss real-time communication method and system based on layered coding, and related devices |
| US8938663B2 (en) * | 2012-01-24 | 2015-01-20 | Broadcom Corporation | Modem architecture for joint source channel decoding |
-
2016
- 2016-03-16 CN CN201610148403.XA patent/CN105812097B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436990A (en) * | 2008-12-23 | 2009-05-20 | 深圳华为通信技术有限公司 | Method for automatically adjusting encoding rate, receiving device and communication system |
| US8938663B2 (en) * | 2012-01-24 | 2015-01-20 | Broadcom Corporation | Modem architecture for joint source channel decoding |
| CN104009819A (en) * | 2013-02-22 | 2014-08-27 | 中兴通讯股份有限公司 | Anti-packet-loss real-time communication method and system based on layered coding, and related devices |
| CN103888846A (en) * | 2014-03-04 | 2014-06-25 | 浙江大学 | Wireless video streaming service self-adaption rate control method based on QoE |
Non-Patent Citations (1)
| Title |
|---|
| "Sync & Sense Enabled Adaptive Packetization VoIP";Boonchai Ngamwongwattana;《Diss.University of Pittsburgh》;20071231;第76-94页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105812097A (en) | 2016-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105812097B (en) | An adaptive adjustment method of AMR coding rate based on network state | |
| KR102797095B1 (en) | Audio coding method and device | |
| KR101353847B1 (en) | Method and apparatus for detecting and suppressing echo in packet networks | |
| US8102872B2 (en) | Method for discontinuous transmission and accurate reproduction of background noise information | |
| KR100575193B1 (en) | Decoding Method and System Including Adaptive Postfilter | |
| CN101854308B (en) | Self-adaptation realizing method of high-tone quality service network of VoIP system | |
| CN102543090B (en) | Code rate automatic control system applicable to variable bit rate voice and audio coding | |
| CN111681664A (en) | Method, system, storage medium and equipment for reducing audio coding rate | |
| CN101826891B (en) | Enhanced echo cancellation | |
| CN117793078B (en) | Audio data processing method and device, electronic equipment and storage medium | |
| CN101636786B (en) | Method of transmitting data in a communication system | |
| CN102800318A (en) | Device and method for transmitting and receiving audio data streams | |
| CN101120401B (en) | Sound source supply device and sound source supply method | |
| CA2442044A1 (en) | Method and apparatus for controlling an operative setting of a communications link | |
| US20110004471A1 (en) | Method and means for encoding background noise information | |
| CN110010141A (en) | Method and apparatus for the DTX hangover in audio coding | |
| CN101170590B (en) | A method, system and device for background noise coded stream transmission | |
| CN112002338A (en) | Method and system for optimizing audio coding quantization times | |
| Beritelli et al. | Hybrid multimode/multirate CS-ACELP speech coding for adaptive voice over IP | |
| Sasaki et al. | Voice activity detection and transmission error control for digital cordless telephone system | |
| Norlund | Voice over IP in PDC Packet Data Network | |
| Thorpe et al. | Changes in voice quality judgments as a function of background noise level in the listening environment | |
| CN119030959A (en) | A method for optimizing outbound voice communication quality based on intelligent AI | |
| Leite | Characterisation of noisy speech channels in 2G and 3G mobile networks | |
| Neves | Methods for quality enhancement of voice communications over erasure channels |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |