CN102710374B - Speed control method in wireless streaming media transmission - Google Patents

Speed control method in wireless streaming media transmission Download PDF

Info

Publication number
CN102710374B
CN102710374B CN201210170301.XA CN201210170301A CN102710374B CN 102710374 B CN102710374 B CN 102710374B CN 201210170301 A CN201210170301 A CN 201210170301A CN 102710374 B CN102710374 B CN 102710374B
Authority
CN
China
Prior art keywords
packet loss
rate
speed
channel
time interval
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
Application number
CN201210170301.XA
Other languages
Chinese (zh)
Other versions
CN102710374A (en
Inventor
杨晋生
胡自胜
陈为刚
王蓉
张菊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin University
Original Assignee
Tianjin University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tianjin University filed Critical Tianjin University
Priority to CN201210170301.XA priority Critical patent/CN102710374B/en
Publication of CN102710374A publication Critical patent/CN102710374A/en
Application granted granted Critical
Publication of CN102710374B publication Critical patent/CN102710374B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention relates to wireless multimedia communication. In order to alleviate transmission quality shaking caused by an unreliable wireless channel when streaming media is transmitted in the wireless environment, further improve the transmission quality and increase the utilization ratio of the wireless channel. The technical scheme adopted by the invention is as follows: compared with a traditional speed control method, a speed control method in wireless streaming media transmission is characterized by comprising the following steps: obtaining the packet loss rate and business data passing rate parameters by smoothing treatment in a shorter interval; adjusting the streaming media transmission speed based on the parameters and an improved additive increase and multiplicative decrease algorithm, when the channel is in underload, determining the speed amplification by a variable constant additive factor and a constant additive factor determined by the packet loss, and when the channel is in overload, determining the speed reduction range by a variable constant multiplicative factor and a constant reductive factor determined by the packet loss; and setting a speed switching threshold, regarding the speed switching to be effective only when the difference between the current predicting speed and the previous predicting speed exceeds the set threshold. The speed control method in the wireless streaming media transmission is mainly applied to wireless multimedia communication.

Description

Method of rate control in wireless flow media transmission
Technical field
The invention belongs to wireless multimedia communication technical field, relate to a kind of increasing that adds based on packet loss, business datum percent of pass and improvement to take advantage of and subtract the method for rate control that (AIMD) algorithm ensures wireless flow media transmission service quality, specifically, the space problem solver based on constraint satisfaction framework is related to.
Background technology
In recent years, along with the fast development of the communication technology, network technology, video coding and decoding technology, the service content of network service becomes more rich and varied.Multimedia communication in conjunction with information transmission such as voice, video, data has become the inexorable trend of communication service development, wherein multimedia video communication due to have continuity transmission, data volume large, to real-time, reliability requirement high and become study hotspot.
Wired multimedia communication technology has become ripe, as the technology such as visual telephone, video request program (VoD:Video on Demand), rich email, cable digital TV and video monitoring based on H.261/H.263 standard obtains extensive use.But, although wire message way can provide more stable reliable transmission environment for multimedia communication, the intrinsic drawback of wire transmission, as more high in having higher requirement, cost to wiring environment, limit further developing of multimedia communication.Multimedia communication service based on wireless transmission receives the concern of people.
The development of nearly ten years Radio Transmission Technologys is advanced by leaps and bounds, and the transmission bandwidth that wireless network can provide have also been obtained lifting at full speed.G mobile Communication network can provide the speed of 2Mbps for fixed position service, and under the state of people's walking, speed can reach 384Kbps, also can reach 128Kbps, reached the bandwidth requirement of multimedia video transport service in the speed of a motor vehicle is advanced.In forth generation (4G) wireless telecommunication system, transmission rate can reach more than 80Mbps; IEEE 802.11 series standard can provide the transmission rate of every number of seconds million to hundreds of million; Physical layer transmission rate has been risen to Gbps by the ultrahigh speed WLAN standard with independent intellectual property right that China proposes in the recent period especially.The quick growth of wireless communication bandwidth, for multimedia communication provides better opportunity to develop.But it is high with the characteristic become in time that wireless channel has error rate, the unreliable meeting of wireless channel causes multimedia video quality that serious shake occurs, thus makes to ensure wireless flow media transmission service quality faces enormous challenge.
For these reasons, there has been proposed multiple method of rate control to improve wireless flow media transmission service quality, improve its efficiency of transmission.These Streaming Media method of rate control are widely used in the stream media transmission system based on RTP (RTP).RTP (RTP) is a kind of host-host protocol for multimedia data stream on the Internet.Real-time Transport Protocol is made up of two parts: RTP and RTCP (RTCP Real-time Transport Control Protocol).Real-time Transport Protocol is positioned on transport layer protocol TCP and udp protocol, and in order to ensure the real-time transmitted, after Real-time Transport Protocol adds RTP packet header before load, just direct bag by RTP consigns to UDP and transmit.RTCP mainly adds up the various transformation parameters (packet loss, time delay etc.) in transmitting procedure.Traditional rate control algorithm is all that the transformation parameter utilizing RTCP to provide carries out speed control.According to the speed control principle of current Streaming Media rate control algorithm, two classes can be divided into:
■ adds increasing to be taken advantage of and subtracts (AIMD) algorithm.AIMD is in TCP/IP mode, for network layer provides the algorithm of congestion control.This algorithm can provide reliable data transport service for TCP/IP network.It is the state judging network by comparing actual measurement packet loss and the packet loss threshold value preset.If packet loss is greater than packet loss thresholding, illustrate that generation is congested, then reduce the factor according to the constant multiplicative preset and reduce transmission rate; Otherwise, increase transmission rate according to the constant additivity growth factor preset.Do not consider the relation of the degree that present channel is congested and speed adjusting range when AIMD algorithm carries out speed adjustment, be a kind of linear algorithm, calculate very simple.But, use AIMD algorithm easily to make transmission rate generation zigzag fluctuate, be unfavorable for improving transmission service quality.
■ TCP is friendly, and speed controls (TFRC, TCP Friendly Rate Control) algorithm.TFRC algorithm carries out the adjustment of transmission rate based on TCP throughput model, as shown in formula (1)
SendRate ( λ ) = 1.5 × 2 3 × MTU RTT × λ ≈ 1.22 × MTU RTT × λ - - - ( 1 )
Wherein MTU is MTU; λ is packet loss, and RTT is Loop Round Trip Time, and SendRate is network throughput; This algorithm can obtain more stable and level and smooth transmission rate, but this algorithm maximizes according to Internet usage to carry out speed adjustment, have ignored the perceived quality of video flowing.
Above-mentioned two class algorithms there is common feature: two class algorithms are all that the transformation parameter statistical information that utilizes RR in rtcp protocol (Receiver Report) to wrap to provide is to carry out the control of streaming media speed.Because in rtcp protocol, RR bag is that interval 5s sends, this just means that stream media transmission system often can only carry out a transmission rate control by 5s, and this is enough for the metastable wire transmission of channel; But bandwidth being had to the wireless transmission of time-varying characteristics, the control of transmission rate cannot the change of adaptive channel.
Summary of the invention
The present invention is intended to overcome the deficiencies in the prior art, the method of rate control of streaming media under a kind of wireless environment is provided, to reduce unreliable due to wireless channel when Streaming Media transmits in the wireless context and transmission quality that is that cause is shaken, and then to improve the utilance of streaming media service quality and wireless channel.For achieving the above object, the technical scheme that the present invention takes is, the method for rate control in wireless flow media transmission, obtains packet loss and business datum percent of pass parameter by smoothing processing in the time interval shorter than conventional rate control method; The adjustment streaming media speed that cuts algorithm is taken advantage of in the increasing that adds based on above-mentioned parameter and improvement, when channel underloading, the change constant add factor determined by packet loss and constant add factor determine speed increasing degree, when channel overloads, the change constant multiplicative Summing Factor constant subtractive factor determined by packet loss determines rate reduction amplitude; Setting speed handoff threshold, only have when this predicted velocity and last time predicted velocity difference exceed and set thresholding time just think that speed switches effectively.
Comprise the following steps:
(1) obtaining packet loss and business datum percent of pass: at the receiving terminal of stream media transmission system, obtaining packet loss and business datum percent of pass parameter by carrying out statistical analysis to the flow media data packet received;
(2) new transmission rate is predicted: after the receiving terminal of stream media transmission system obtains packet loss and business datum percent of pass, based on the AIMD algorithm improved, analyze wireless channel state at this moment and predict the new transmission rate being applicable to now wireless signal-path band width;
(3) judge whether to carry out transmission rate adjustment: after the receiving terminal of stream media transmission system predicts the transmission rate made new advances, need to compare this and predict whether the difference between the new speed that new speed and last time are predicted exceedes the threshold value of setting, if exceed this threshold value, then the adjustment of this subrate effectively, otherwise the adjustment of this subrate is invalid.
Obtain packet loss and business datum percent of pass step as follows:
1) according to radio channel state, the time interval Δ t that packet loss and business datum percent of pass obtain is determined;
2) participate in the time interval number N calculated when determining acquisition i-th time interval parameter, and participate in the i-th-j time interval corresponding weights of calculating;
3) statistic fluid media transport parameters, while receiving terminal receiving stream media packet, adds up i-th time interval intrinsic parameter, comprises number of dropped packets LPN i, total transmission bag number SPN i, business datum amount D iand accurately interval time T i;
4) packet loss P in i-th time interval is calculated iwith business datum percent of pass R i, by being weighted smoothing computation packet loss and business datum percent of pass to the i-th-N+1 the time interval to i-th transformation parameter that the time interval obtains.Obtain time interval Δ t in packet loss and business datum percent of pass step and get 2s, the time interval number N value participating in calculating is 2 or 3.
The AIMD algorithm improved comprises following steps:
1) according to the relation of test statistics Video transmission system receiving terminal video quality and packet loss, in set algorithm, packet loss upper-lower door limit value is respectively P loss_maxand P loss_min, maximum traffic data rate R maxand minimal service data rate R min;
2) actual measurement packet loss P is compared iwith default packet loss thresholding magnitude relationship, determine that channel status is overload, underloading or load balance;
3) when channel overloads, i.e. P i>=P loss_max, calculate the difference factor m=P of channel overload i-P loss_maxand overload rate reduction factor-beta k=β k-1× (1-0.6 × m) 2, wherein k represents the number of times of overload continuously, according to formula R new=MAX{ β k× R i-50, R mincalculate new transmission rate, wherein R ibusiness datum percent of pass in i-th time interval, R minit is the minimal service data rate of default;
4) when channel underloading, i.e. P i≤ P loss_min, calculate the difference factor m=λ × (P of channel underloading loss_min-P i) and speed recruitment α=0.2e m× R 0+ 60, wherein R 0for the speed growth factor preset; According to formula R new=MIN{ α+R i, R maxcalculate new transmission rate, wherein R ibusiness datum percent of pass in i-th time interval, R maxit is the maximum traffic data rate that system allows;
5) when channel loading balances, without the need to adjusting speed, rate of withdrawal control method.
The AIMD algorithm realization step 3 improved) in overload first rate reduction factor-beta 0span be [0.85,0.95]; The AIMD algorithm realization step 4 improved) medium-rate increase factor R 0span is [80,150].
Speed handoff threshold is set, threshold value R in method of rate control thgenerally be set as 40 ~ 80Kb/s.
Technical characterstic of the present invention and effect:
The present invention adopts less channel status acquisition time interval to analyze channel status, makes the method can reflect the change of wireless channel faster.The present invention is based on the transmission rate that the AIMD algorithm predicts of improvement is suitable, relative to traditional AIMD algorithm, modified hydrothermal process increases the impact of packet loss on transmission rate adjusting range, and the prediction of speed can be made more close to wireless signal-path band width.The present invention has taken into full account the characteristic of wireless channel, can as a kind of effective ways improving stream media transmission quality in the wireless context.
Beneficial effect of the present invention is also:
■ reflects that wireless channel changes fast.Wireless channel have decline, time the characteristic such as change, the change of channel status is relatively very fast.The present invention is based on acquisition packet loss of the self-defined less time interval and business datum percent of pass, smoothing processing has been carried out to important parameter simultaneously, met the time-varying characteristics of wireless channel, made the adjustment of speed more timely;
■ reduces transmission rate shake.The AIMD rate adaptation adjustment algorithm of the present invention by improving, takes full advantage of packet loss and the important parameter of business datum percent of pass these two adjusts transmission rate, speed is adjusted more accurate, decreases zigzag fluctuation;
■ reduces overhead.The present invention, by adding judgment mechanism before transmission rate adjustment feedback, can reduce the expense that system is unnecessary;
■ improves wireless flow media transmission service quality and radio channel utilization.The present invention is controlled by streaming media transmitting procedure medium-rate, reduces the shake of packet loss and transmission rate, thus improves transmission service quality and channel utilization.
Accompanying drawing explanation
Fig. 1 method of rate control overview flow chart.
Fig. 2 packet loss and business datum percent of pass acquisition methods flow chart.
The AIMD algorithm general flow chart that Fig. 3 improves.
Speed adjustment flow chart under channel overload condition in the AIMD algorithm that Fig. 4 improves.
Speed adjustment flow chart in channel underloading situation in the AIMD algorithm that Fig. 5 improves.
The structure composition schematic diagram of Fig. 6 wireless real-time video transmission system.
Fig. 7 embodiment system performance testing result Y-PSNR figure.
Fig. 8 embodiment system performance testing result packet loss and business datum percent of pass figure.
Embodiment
Based on above consideration, the present invention is directed to and improve wireless flow media transmission service quality and this problem of channel utilization, propose the method for rate control of a kind of wireless flow media transmission.The method is based on stream media transmission system, the parameters such as packet loss and business datum percent of pass are added up with the time interval shorter compared with conventional rate control method at system receiving terminal, based on above-mentioned two parameter, by the AIMD algorithm predicts streaming media service message transmission rate improved, switch effectively in judgement speed and feed back to transmitting terminal by Transmission Control Protocol.Compare with traditional Streaming Media method of rate control, the present invention can make the bandwidth of streaming media speed adaptive wireless channel more accurately, thus ensures streaming media service quality, improves radio channel utilization.
In order to achieve the above object, the invention provides the method for rate control in the transmission of a kind of wireless flow media, it is characterized in that: the method obtains the parameter such as packet loss and business datum percent of pass by smoothing processing within the time interval shorter than conventional rate control method; The adjustment streaming media speed that cuts algorithm is taken advantage of in the increasing that adds based on above-mentioned parameter and improvement, when channel underloading, the change constant add factor determined by packet loss and constant add factor determine speed increasing degree, when channel overloads, the change constant multiplicative Summing Factor constant subtractive factor determined by packet loss determines rate reduction amplitude; Last the present invention sets speed handoff threshold, only have when this predicted velocity and last time predicted velocity difference exceed and set thresholding time just think that speed switches effectively.
The invention provides the method for rate control in the transmission of a kind of wireless flow media, it is characterized in that: said method comprising the steps of:
(1) packet loss and business datum percent of pass is obtained.At the receiving terminal of stream media transmission system, obtain the parameter such as packet loss and business datum percent of pass by carrying out statistical analysis to the flow media data packet received;
(2) new transmission rate is predicted.After the receiving terminal of stream media transmission system obtains packet loss and business datum percent of pass, based on the AIMD algorithm improved, analyze wireless channel state at this moment and predict the new transmission rate being applicable to now wireless signal-path band width;
(3) judge whether to carry out transmission rate adjustment.After the receiving terminal of stream media transmission system predicts the transmission rate made new advances, need to compare this and predict whether the difference between the new speed that new speed and last time are predicted exceedes the threshold value of setting, if exceed this threshold value, then the adjustment of this subrate effectively, otherwise the adjustment of this subrate is invalid.
The invention provides a kind of wireless flow media transmission in method of rate control, it is characterized in that: above-mentioned acquisition packet loss and business datum percent of pass step as follows:
1) according to radio channel state, the time interval Δ t that packet loss and business datum percent of pass obtain is determined;
2) participate in the time interval number N calculated when determining acquisition i-th time interval packet loss and business datum percent of pass, and participate in the i-th-j time interval corresponding weights α of calculating jand μ j, wherein 0≤j≤N;
3) statistic fluid media transport parameters.While receiving terminal receiving stream media packet, add up i-th time interval intrinsic parameter, comprise number of dropped packets LPN i, total transmission bag number SPN i, business datum amount D iand accurately interval time T i;
4) packet loss and the business datum percent of pass in i-th time interval is calculated.By being weighted smoothing computation packet loss and business datum percent of pass to the i-th-N+1 the time interval to i-th transformation parameter that the time interval obtains.
Method of rate control in wireless flow media transmission proposed by the invention, is characterized in that: obtain time interval Δ t in packet loss and business datum percent of pass step and generally get about 2s, the time interval number N value participating in calculating is 2 or 3.
Method of rate control in wireless flow media transmission proposed by the invention, is characterized in that: the AIMD algorithm of improvement comprises following steps:
1) according to the relation of test statistics Video transmission system receiving terminal video quality and packet loss, in set algorithm, packet loss upper-lower door limit value is respectively P loss_maxand P loss_min, maximum traffic data rate R maxand minimal service data rate R min;
2) actual measurement packet loss P is compared iwith default packet loss thresholding magnitude relationship, determine that channel status is overload, underloading or load balance;
3) when channel overloads, i.e. P i>=P loss_max, calculate the difference factor m=P of channel overload i-P loss_maxand overload rate reduction factor-beta k=β k-1× (1-0.6 × m) 2, wherein k represents the number of times of overload continuously, according to formula R new=MAX{ β k× R i-50, R mincalculate new transmission rate, wherein R ibusiness datum percent of pass in i-th time interval, R minit is the minimal service data rate of default;
4) when channel underloading, i.e. P i≤ P loss_min, calculate the difference factor m=λ × (P of channel underloading oss_min-P i) and speed recruitment α=0.2e m× R 0+ 60, wherein R 0for the speed growth factor preset; According to formula R new=MIN{ α+R i, R maxcalculate new transmission rate, wherein R ibusiness datum percent of pass in i-th time interval, R maxit is the maximum traffic data rate that system allows;
5) when channel loading balances, without the need to adjusting speed, rate of withdrawal control method.
Method of rate control in wireless flow media transmission proposed by the invention, is characterized in that: the AIMD algorithm realization step 3 of improvement) in overload first rate reduction factor-beta 0span be [0.85,0.95]; The AIMD algorithm realization step 4 improved) medium-rate increase factor R 0span is [80,150].
Method of rate control in wireless flow media transmission proposed by the invention, is characterized in that: set speed handoff threshold in method of rate control, threshold value R thgenerally be set as 40 ~ 80Kb/s.
In order to make the object of the invention, technical scheme feature and advantage clearly, describe the present invention in detail below in conjunction with accompanying drawing.
The present invention is a kind of method of rate control for streaming media under wireless channel environment.Streaming media, particularly real-time video transmission, have continuity transmission, data volume large, to real-time, reliability requirement high, if channel is unreliable or occur congested situation, the service quality of streaming media can be had a strong impact on.In case of a wired network, the main cause affecting stream media transmission quality is the congested of network.But in the wireless context, the principal element affecting stream media transmission quality is bandwidth time variation and the unreliability of wireless channel.Therefore the present invention is mainly for the streaming media under wireless environment, by analyzing the transmission situation of a period of time thus judging the state of wireless channel, finally dopes suitable transmission rate.The present invention can reduce the packet loss in streaming media process, thus improves service quality and the radio channel utilization of wireless flow media transmission.
The present invention adopts the time interval less than conventional flow media method of rate control to obtain packet loss and business datum percent of pass, with the transmission rate that the AIMD algorithm algorithm predicts improved is new.This makes the method can reflect the change of wireless channel fast, the new transmission rate of Accurate Prediction, thus improves the service quality of wireless flow media transmission and the utilance of wireless channel.
Before use the present invention, need to determine parameter preset of the present invention according to the Stream Media Application sight of reality and wireless environment, the performance of these parameter values on method of rate control provided by the invention has larger impact.
■ maximum traffic data rate R maxand minimal service data rate R mim.These two parameters are maximum and the minimum value of traffic data rate adjustment respectively.Their value determines the actual requirement of business datum throughput according to streaming media service.
■ maximum packet loss thresholding P loss_maxand minimum packet loss thresholding P loss_min.Maximum packet loss thresholding represents the degrees of tolerance of stream media transmission system to packet loss, and minimum packet loss thresholding represents the degrees of tolerance of stream media transmission system to channel underloading.
The time interval Δ t that ■ speed controls.This value represents in the process of streaming media, and every the Δ t time, rate control algorithm streaming media transmission rate according to the present invention adjusts.If this value value is too large, then cannot reflect the time variation of wireless channel; If this value value is too little, then can lead speed shake larger.
The following detailed description of each step of the inventive method.
Fig. 1 is the inventive method overview flow chart.As can be seen from the figure, the inventive method mainly comprises that packet loss and business datum percent of pass obtain, the AIMD algorithm speed adjustment that improves and system velocity be switched and determined three parts.Based on the time interval Δ t setting speed control, the inventive method obtains streaming media parameter within the Δ t time period, use the AIMD algorithm improved to carry out the prediction of new transmission rate in the Δ t last moment, the validity that system velocity switches is judged simultaneously.Method of rate control of the present invention always with Δ t for time interval circular flow.In order to make the detailed process introduced clearly below, be all described for i-th time interval.
Detailed process of the present invention is as follows:
(1) packet loss and business datum percent of pass.Receiving terminal is while receiving stream media packet, (the packet number that transmitting terminal sends altogether is comprised to the transmission information of flow media data packet in i-th time interval, the packet number that receiving terminal receives, receiving terminal receives total business datum amount and correct time spacing value) add up, thus calculate packet loss P in i-th time interval iwith business datum percent of pass R i;
(2) new transmission rate is predicted.Receiving terminal acquisition packet loss and business datum percent of pass after, by packet loss P icompare with the packet loss thresholding preset, thus know current channel status: overload, underloading and balance.According to three kinds of channel statuss and packet loss, business datum percent of pass, the AIMD algorithm of improvement can accurately predict new speed;
(3) judge whether to carry out transmission rate adjustment.When obtaining new transmission rate R newafter, by this value and last time rate prediction value R new_lastcompare, only have and work as | R new-R new_last| > R thtime, just notification streams media system transmitting terminal changes transmission rate, wherein R thfor the speed adjustment threshold value preset.
The handling process of packet loss and business datum percent of pass is obtained as shown in Figure 2 in Fig. 1, receiving terminal obtains the information such as frame of video belonging to the sequence number of current data packet, current data packet and bag load data length by the flow media data packet that receives, thus calculates the parameter required for rate adaptation adjustment.
The flow process that packet loss and business datum percent of pass obtain is as follows:
(1) first initialization counter.Comprise initializes video frame number counter F_cnt, receiving stream media data packet count device P_cnt and business datum batching counter D i.Because video requency frame data amount is larger, transmit so transmitting terminal can be cut into several packet.Therefore often receive a flow media data packet, whether this packet is the initial data bag of a frame of video first to need the flag bit preset according to transmitting terminal to judge, if it is frame counter F_cnt is added 1, otherwise remains unchanged; Then obtain the payload traffic data length of this packet, this length is added to business datum batching counter D ion; Last package counting facility P_cnt is corresponding adds 1.
(2) determine to count initial parameters.When receiving stream media data packet count device P_cnt is 1, the channel estimating amount of a new time period obtains and starts, obtain sequence number psn1 (in order to check streaming media whether packet loss, needing for each packet adds 16 bit sequence) and the current time time1 of this packet;
(3) determine to count terminal parameter.In order to adapt to the Rapid Variable Design of wireless channel, in this system, setting is when frames received reaches 30 frame, and this time period terminates, and obtains sequence number psn2 and the current time time2 of now packet.The video frame rate of system is 15 frames/s, and therefore each time interval is approximately Δ t=2s;
(4) the exact value T in this time interval is calculated i=time2-time1, and at T ithe overall budget number SPN that in time, transmitting terminal sends i=psn2-psn1+1, number of dropped packets LPN i=SPN i-P_cnt.Because in data packet head, bag sequence number accounts for 16 bits, therefore maximum sequence number is 65535, there will be sequence number overflow when system cloud gray model, namely occurs sending bag number SPN i< 0 situation, now SPN i=psn2-psn1+1+65536.
(5) packet loss P is calculated i.Because predetermined period Δ t is less, and packet loss reflection is the congestion state of channel, and with moment channel congestion state is relevant before, the weighted value of therefore getting various packet statistics result in the top n time interval calculates, and N value depends on channel circumstance.Calculate packet loss according to formula (2), in formula as N=3, weighted value α value is: α 0=0.5, α 1=0.3, α 2=0.2; As N=2, weighted value α value is: α 0=0.7, α 1=0.3; LPN i-jrepresent in the i-th-j the time interval packet loss sum added up and obtain, SPN i-jrepresent that in the i-th-j the time interval, transmitting terminal sends packet sum.
P i = &Sigma; j = 0 N - 1 &alpha; j &times; LPN i - j &Sigma; j = 0 N - 1 &alpha; j &times; SPN i - j , &Sigma; j = 0 N - 1 &alpha; j = 1 - - - ( 2 )
(6) computing service data pass rate R i.According to formula (3) computing service data pass rate, in formula as N=3, weighted value μ 0=0.7, μ 1=0.2, μ 2=0.1; As N=2, weighted value μ 0=0.8, μ 1=0.2; D i-jrepresent the business datum amount received in the i-th-j the time interval, T i-jit is the correct time value in the i-th-j the time interval;
P i = &Sigma; j = 0 N - 1 &mu; j &times; D i - j &Sigma; j = 0 N - 1 &mu; j &times; T i - j , &Sigma; j = 0 N - 1 &mu; j = 1 - - - ( 3 )
(7) reset to the various counters of above-mentioned use, the channel estimating gauge starting following time interval is calculated.
The AIMD algorithm general flow chart improved in Fig. 1 as shown in Figure 3.Traditional AIMD algorithm does not consider the impact of real-time packet loss parameter, present rate is multiplied by reduce transmission rate with invariant when channel overloads, add that current transmission rate is to increase transmission rate with constant during channel underloading, the zigzag of transmission rate can be caused like this to fluctuate.The present invention considers packet loss this important parameter when adjusting speed, and packet loss is larger with threshold value difference, and the amplitude that speed adjusts is larger.
The AIMD algorithm steps improved is as follows:
(1) the maximum packet loss thresholding P that stream media transmission system allows is set loss_maxand minimum packet loss thresholding P loss_min;
(2) by packet loss P icompare with the packet loss thresholding preset.If packet loss P ibe greater than maximum packet loss thresholding P loss_max, then present channel overload is described, needs changing down to improve transmission of video service quality;
(3) if packet loss P ibe less than minimum packet loss thresholding P loss_min, then present channel underloading is described, needs to advance the speed to improve radio channel utilization;
(4) if packet loss P ibe greater than minimum packet loss thresholding P loss_minand be less than maximum packet loss thresholding P loss_max, then channel loading balance is described, without the need to carrying out speed adjustment, rate of withdrawal control algolithm.
As shown in Figure 4, the AIMD algorithm of improvement determines new transmission rate by the degree and channel current service data percent of pass considering channel overload to handling process in Fig. 2 during channel overload.The stylish speed determination flow process that overloads is as follows:
(1) first initialization channel overload first rate reduction factor-beta 0, this value affects the amplitude of rate reduction, and value is 0.93 in the present invention;
(2) the number of times k that channel overloads continuously is judged.When channel overloads continuously, illustrate that channel enters larger decline, need the amplitude reduced of relatively advancing the speed.
(3) the difference factor m=P of this secondary channel overload is calculated i-P loss_max, wherein P ifor actual measurement packet loss, P loss_maxfor the maximum packet loss of systemic presupposition;
(4) this secondary channel overload rate reduction factor-beta is calculated k=β k-1× (1-0.6 × m) 2;
(5) the new speed R that the adjustment of this rate adaptation produces is calculated new=MAX{ β k× R i-50, R min, wherein R ifor business datum percent of pass, R minfor systemic presupposition minimal service message transmission rate;
As shown in Figure 5, rate adaptation control module determines new transmission rate by the degree and channel current service data percent of pass considering channel underloading to handling process in Fig. 2 when channel underloading.Underloading stylish speed determination flow process is as follows:
(1) initializing rate increases factor R 0, this value affects the amplitude that speed increases, and value is 100 in systems in which;
(2) regeneration rate reduces factor-beta 0, for lower secondary channel overload is prepared;
(3) the packet loss difference factor m=λ × (P of this secondary channel underloading is calculated loss_min-P i), wherein λ is packet loss difference multiplication factor;
(4) speed recruitment α=0.2e under this secondary channel underloading is calculated m× R 0+ 60, wherein R 0speed increases the factor;
(5) the new speed R that the adjustment of this rate adaptation produces is calculated new=MIN{ α+R i, R max, wherein R ibe business datum percent of pass in i-th time interval, R maxfor maximum traffic data rate;
Wireless flow media method of rate control provided by the invention can be used for several scenes, comprises wireless video monitoring, wireless video conference, wireless videophone and Streaming Media downloading and playing etc.Method provided by the invention effectively can improve the service quality of these application.
Apply as embodiment of the present invention using the real-time video higher to quality of service requirement below.Streaming Media method of rate control is used in end-to-end wireless real-time video transmission system, thus ensures the service quality of real-time video in transmitting procedure.
Fig. 6 is the system block diagram of the concrete real-time proposals of the present invention, and this system is the end-to-end wireless video RTTS realized based on embedded Linux platform.This system is using the Ad-Hoc of 802.11b/g standard (wireless self-organization network) pattern as the wireless environment of real-time video transmission.Reach 54Mb/s because the most high energy of 802.11b/g standard physical layer speed gets at, the bandwidth requirement of real-time video transmission can be reached.This system is made up of video sending end and video receiver, and video sending end mainly comprises video acquisition coding module and video data sends module, and video receiver mainly comprises data reception module, video decode display module and Rate control module.Below this system modules and feature thereof are described in detail.
Video sending end:
(1) video acquisition coding module.This module obtains by the parameter such as frame per second, resolution configuring camera video acquisition the real-time video met the demands, and in order to reduce the bandwidth needed for transmission, carries out H264 video format coding by coding module to it.During system cloud gray model, the major parameter of capturing and coding module is as follows:
■ video resolution: 640x400, remains unchanged in transmitting procedure;
■ video frame rate: 15 frames/s, remains unchanged in transmitting procedure;
■ encoder outgoing traffic data rate: be initialized as 2048Kb/s, adjusted by Rate control module afterwards;
(2) video data sends module.In order to ensure the real-time Transmission of video data, we adopt RTP as the host-host protocol of video data stream.Use use the present invention that RTP can be convenient as video data transmission agreement, because information entrained by RTP packet header when Real-time Transport Protocol defines stream media information transmission, comprise the source of this RTP packet, the sequence number of packet and timestamp etc.According to the source of RTP packet, stream media transmission system can receive multi-path media source at receiving terminal simultaneously; According to the sequence number of RTP packet, the packet loss of stream media transmission system within some time periods can be calculated.By obtaining the length of RTP data packet traffic data, stream media transmission system business datum percent of pass in a period of time can be calculated.One-frame video data is carried out cutting packing according to RTP in this module by system, then consigns to UDP and transmits.
Video receiver:
(1) data reception module.This functions of modules receives RTP packet, and then unpack RTP packet, and being combined by the RTP packet load of same frame becomes frame of video.
(2) video decode display module.The RTP Data Synthesis received by after the frame of video compressed, is carried out the decoding of H264 video format by decoder module to it by data reception module.Because decoded video data is YUV420 form, shown by LCD after therefore needing to be converted into rgb format.
(3) Rate control module.This module is method provided by the invention embodiment in this embodiment, is the important module ensureing this system real-time video transmission service quality.This module forms primarily of 3 submodules, comprises packet loss and business datum percent of pass and obtains submodule, the AIMD speed adjustment submodule of improvement and system velocity and be switched and determined submodule.Being described below of three submodules:
■ packet loss and business datum percent of pass obtain submodule.This submodule, while data reception module carries out RTP receives data packets, is added up transformation parameter according to method provided by the invention, thus obtains packet loss and business datum percent of pass.The cycle set that above-mentioned parameter obtains by system is the time that receiving terminal receives 30 frame videos, and because system video frame per second is that 15 frames/s remains unchanged, therefore the time is about 2s.When this submodule calculates packet loss and business datum percent of pass according to formula provided by the invention (2) and (3), the value of parameter is as follows:
participate in the time interval number N=3 calculated;
packet loss smoothing factor α 0=0.5, α 1=0.3, α 2=0.2;
μ 0＝0.7，μ 1＝0.2，μ 2＝0.1。
The AIMD speed adjustment submodule that ■ improves.This submodule is after acquisition packet loss and business datum percent of pass, and adjust transmission rate according to the AIMD algorithm of improvement provided by the invention, the value of various parameter is as follows:
packet loss upper-lower door limit value, P loss_max=0.03, P loss_min=0.008;
maximum traffic data rate R max=2048kb/s, minimal service data rate R min=256kb/s;
the rate reduction of channel overload first factor-beta 0=0.93;
speed increases factor R 0=100.
■ system velocity is switched and determined submodule.This module mainly increases overhead in order to anti-locking system occurs frequent switching.According to system velocity switching determination method provided by the invention, its threshold value R judged thbe set as 60Kb/s.
As shown in Figure 6, this system has by two wireless transmission links.Article 1, be the video data transmission link based on RTP/UDP/IP, the transmission of video that primary responsibility is real-time; Article 2 is the FEEDBACK CONTROL link based on TCP/IP, the transmitting of the rate adjustment information of primary responsibility Rate control module.
In order to assess said system performance, carry out the parameters such as the PSNR (Y-PSNR) having added up packet loss, business datum percent of pass and transmission of video in video transmitting procedure in this system.The test environment of systematic function is the corridor having many different radio access points, and between different wireless signal, interference ratio is comparatively large, and transmitting terminal system and receiving terminal system are at a distance of 42m.The result of performance test is as follows.
As shown in Figure 7, be the PSNR value of carrying out H264 decoded video audio data before computing system transmitting terminal H264 compresses with receiving terminal.Therefore, the loss of video quality shown in Fig. 7 is made up of two parts: transmission loss and H264 encoding and decoding are lost.
The PSNR value of transmission of video and the evaluation of people to video-aware quality have certain relation, the PSNR value in Fig. 7 are carried out segmentation and add up as follows:
pSNR value is greater than 37, and video level is 5, and video quality is very good, accounts for sum 88.312%;
pSNR value be positioned at interval (31,37], video level is 4, and video quality is good, account for sum 9.5238%;
pSNR value be positioned at interval (25,31], video level is 3, and video quality is general, accounts for sum 0%;
pSNR value be positioned at interval (20,25], video level is 2, and video quality is poor, account for sum 0.8658%;
pSNR value is less than 20, and video level is 4, poor video quality, accounts for sum 1.2987%;
As can be seen from the result of statistics, this system can provide higher transmission of video service quality, the video quality that can provide in 97.8358% situation, only has when wireless channel decline is very large, the reduction causing video quality that can be of short duration.
Fig. 8 is the system traffic data rate of statistics and Parameter Map of packet loss when carrying out transmission of video.As can be seen from the figure, within the most of the time, the packet loss of system is all controlled in less than 2%, but due to the time-varying characteristics of channel, at some time, the packet loss of system is larger, now wireless channel enters larger on the downside, by method of rate control of the present invention, traffic data rate can make an immediate response the change of wireless channel, reduces transmission rate, thus ensures lower packet loss.

Claims (4)

1. the method for rate control in wireless flow media transmission, is characterized in that, obtain packet loss and business datum percent of pass parameter in the time interval shorter than conventional rate control method by smoothing processing; The adjustment streaming media speed that cuts algorithm is taken advantage of in the increasing that adds based on above-mentioned parameter and improvement, when channel underloading, the change constant add factor determined by packet loss and constant add factor determine speed increasing degree, when channel overloads, the change constant multiplicative Summing Factor constant subtractive factor determined by packet loss determines rate reduction amplitude; Setting speed handoff threshold, just thinks that speed switches effectively when only having when this prediction transmission rate and predicted that the difference of transmission rate exceeded and sets thresholding last time; Wherein: be specially by smoothing processing acquisition packet loss and business datum percent of pass parameter in the time interval shorter than conventional rate control method:
1) according to radio channel state, the time interval △ t that packet loss and business datum percent of pass obtain is determined;
2) participate in the time interval number N calculated when determining acquisition i-th time interval parameter, and participate in the i-th-j time interval corresponding weights of calculating;
3) statistic fluid media transport parameters, while receiving terminal receiving stream media packet, adds up i-th time interval intrinsic parameter, comprises number of dropped packets LPN i, total transmission bag number SPN i, business datum amount D iand accurately interval time T i;
4) packet loss P in i-th time interval is calculated iwith business datum percent of pass R i, by being weighted smoothing computation packet loss and business datum percent of pass to the i-th-N+1 the time interval to i-th transformation parameter that the time interval obtains;
The AIMD algorithm improved comprises following steps:
1) according to the relation of test statistics Video transmission system receiving terminal video quality and packet loss, in set algorithm, packet loss upper-lower door limit value is respectively P loss_maxand P loss_min, maximum traffic data rate R maxand minimal service data rate R min;
2) actual measurement packet loss P is compared iwith default packet loss thresholding magnitude relationship, determine that channel status is overload, underloading or load balance;
3) when channel overloads, i.e. P i>=P loss_max, calculate the difference factor m=P of channel overload i-P loss_maxand overload rate reduction factor-beta k=β k-1× (1-0.6 × m) 2, wherein k represents the number of times of overload continuously, according to formula R new=MAX{ β k× R i-50, R mincalculate new transmission rate, wherein R ibusiness datum percent of pass in i-th time interval, R minit is the minimal service data rate of default;
4) when channel underloading, i.e. P i≤ P loss_min, calculate the difference factor m=λ × (P of channel underloading loss_min-P i) and speed recruitment α=0.2e m× R 0+ 60, wherein R 0for the speed growth factor preset, λ is packet loss difference multiplication factor; According to formula R new=MIN{ α+R i, R maxcalculate new transmission rate, wherein R ibusiness datum percent of pass in i-th time interval, R maxit is the maximum traffic data rate that system allows;
5) when channel loading balances, without the need to adjusting speed, rate of withdrawal control method.
2. the method for rate control in wireless flow media transmission as claimed in claim 1, is characterized in that, obtain time interval △ t in packet loss and business datum percent of pass step and get 2s, the time interval number N value participating in calculating is 2 or 3.
3. the method for rate control in wireless flow media transmission as claimed in claim 1, is characterized in that, the AIMD algorithm realization step 3 of improvement) in overload first rate reduction factor-beta 0span be [0.85,0.95]; The AIMD algorithm realization step 4 improved) medium-rate increase factor R 0span is [80,150].
4. the method for rate control in wireless flow media transmission as claimed in claim 1, is characterized in that, set speed handoff threshold, threshold value R in method of rate control thgenerally be set as 40 ~ 80Kb/s.
CN201210170301.XA 2012-05-28 2012-05-28 Speed control method in wireless streaming media transmission Expired - Fee Related CN102710374B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201210170301.XA CN102710374B (en) 2012-05-28 2012-05-28 Speed control method in wireless streaming media transmission

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210170301.XA CN102710374B (en) 2012-05-28 2012-05-28 Speed control method in wireless streaming media transmission

Publications (2)

Publication Number Publication Date
CN102710374A CN102710374A (en) 2012-10-03
CN102710374B true CN102710374B (en) 2015-05-20

Family

ID=46902952

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210170301.XA Expired - Fee Related CN102710374B (en) 2012-05-28 2012-05-28 Speed control method in wireless streaming media transmission

Country Status (1)

Country Link
CN (1) CN102710374B (en)

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104301290B (en) * 2013-07-15 2018-11-09 联想(北京)有限公司 A kind of method and electronic equipment of information processing
CN103532664B (en) * 2013-09-27 2016-09-07 昆山创通微电子有限公司 A kind of method of controlling transmission rate and equipment
CN103731389B (en) * 2014-01-06 2017-02-08 东南大学 OFDM signal transmission method and device
CN104243097A (en) * 2014-09-19 2014-12-24 东软集团股份有限公司 Data transmission method and system based on satellite network
CN104243102A (en) * 2014-09-30 2014-12-24 深圳市云之讯网络技术有限公司 Method and device for controlling code rates on basis of network state evaluation
CN104394484A (en) * 2014-11-12 2015-03-04 海信集团有限公司 Wireless live streaming media transmission method
CN104539917A (en) * 2015-02-03 2015-04-22 成都金本华科技股份有限公司 Method for improving definition of video image
EP3253065B1 (en) 2015-03-20 2019-03-06 Huawei Technologies Co., Ltd. Streaming media resource downloading method and apparatus, and terminal device
CN104936215B (en) * 2015-06-17 2019-01-04 福建联迪商用设备有限公司 A kind of wireless Wi-Fi dynamic rate shaping method and system
CN107332732A (en) * 2017-06-26 2017-11-07 迈普通信技术股份有限公司 A kind of method of sampling of message flow, device and routing device
CN109495326B (en) * 2018-12-28 2021-12-14 北京东土科技股份有限公司 Network bandwidth allocation method and system
CN112887225B (en) * 2019-11-29 2023-09-05 中国移动通信集团安徽有限公司 Flow rate control method and system based on SP service in EPC network
CN111726829B (en) * 2020-04-29 2023-05-23 平安科技（深圳）有限公司 Method, device and computer equipment for acquiring data transmission rate
CN111615006B (en) * 2020-05-29 2022-02-01 北京讯众通信技术股份有限公司 Video code conversion transmission control system based on network state self-evaluation
CN113285835B (en) * 2021-05-26 2022-11-18 广东电网有限责任公司 Carrier network link packet loss rate inference method based on service characteristics

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1596528A (en) * 2002-04-23 2005-03-16 日本电气株式会社 Bit rate control method and device
CN101222296A (en) * 2008-01-31 2008-07-16 上海交通大学 Self-adapting transmission method and system in ascending honeycomb video communication
CN101860895A (en) * 2010-06-11 2010-10-13 上海海维工业控制有限公司 Improved AIMD congestion control method
CN102006221A (en) * 2010-09-30 2011-04-06 北京交通大学 Method for optimizing quality of service of stream media

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1596528A (en) * 2002-04-23 2005-03-16 日本电气株式会社 Bit rate control method and device
CN101222296A (en) * 2008-01-31 2008-07-16 上海交通大学 Self-adapting transmission method and system in ascending honeycomb video communication
CN101860895A (en) * 2010-06-11 2010-10-13 上海海维工业控制有限公司 Improved AIMD congestion control method
CN102006221A (en) * 2010-09-30 2011-04-06 北京交通大学 Method for optimizing quality of service of stream media

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Internet 拥塞控制系统在不同源控制算法作用下的资源竞争分析;戴航等;《计算机科学》;20091130;第36卷(第11期);第40-42页 *

Also Published As

Publication number Publication date
CN102710374A (en) 2012-10-03

Similar Documents

Publication Publication Date Title
CN102710374B (en) Speed control method in wireless streaming media transmission
US9538220B2 (en) Video streaming quality of experience degradation control using a video quality metric
EP2119204B1 (en) Method and arrangement for video telephony quality assessment
RU2305908C2 (en) Adaptive method for evaluating multimedia data transmission speed
CN101909060B (en) Qos control method suitable for real-time streaming media transmission of mobile videos
US20130298170A1 (en) Video streaming quality of experience recovery using a video quality metric
US20130290492A1 (en) State management for video streaming quality of experience degradation control and recovery using a video quality metric
EP2888844B1 (en) Device and method for adaptive rate multimedia communications on a wireless network
US8588071B2 (en) Device and method for adaptation of target rate of video signals
CN111107382B (en) Data flow control method, device and system
EP2772010B1 (en) Optimizing video-call quality of service
US8295348B2 (en) Method for controlling moving picture encoding using channel information of wireless networks
EP2888845B1 (en) Device and method for adaptive rate multimedia communications on a wireless network
CN102868666A (en) Implementation method of streaming media quality monitoring report based on user experience interaction
CN101455086A (en) Video error control based on reverse link information
CN101909208A (en) Video wireless transmission control method suitable for CDMA2000
EP2448174A1 (en) An apparatus for controlling data traffic and a method for measuring QoE
AU2021200428B2 (en) System and method for automatic encoder adjustment based on transport data
CN113891172B (en) Adaptive code rate control method based on RTT (round trip time) and suitable for wireless Mesh network
WO2014209493A1 (en) State management for video streaming quality of experience degradation control and recovery using a video quality metric
WO2014209494A1 (en) Video streaming quality of experience degradation control using a video quality metric
WO2015044104A1 (en) Video traffic management
WO2014209495A1 (en) Video streaming quality of experience recovery using a video quality metric
Ma et al. Early packet loss feedback for webrtc-based mobile video telephony over Wi-Fi
Talaat et al. Etfrc: enhanced tfrc for media traffic

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
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20150520

CF01 Termination of patent right due to non-payment of annual fee