CN101938341A - Cross-node controlled online video stream selective retransmission method - Google Patents
Cross-node controlled online video stream selective retransmission method Download PDFInfo
- Publication number
- CN101938341A CN101938341A CN 201010284648 CN201010284648A CN101938341A CN 101938341 A CN101938341 A CN 101938341A CN 201010284648 CN201010284648 CN 201010284648 CN 201010284648 A CN201010284648 A CN 201010284648A CN 101938341 A CN101938341 A CN 101938341A
- Authority
- CN
- China
- Prior art keywords
- frame
- bag
- importance
- radio link
- link unit
- 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.)
- Granted
Links
Images
Abstract
The invention relates to a cross-node controlled online video stream selective retransmission method comprising the following steps of: when an online-coded single video stream enters a transmitting end, dividing the online-coded single video stream into DFS (Delay Constraint Frame Sets) one by one according to starting and transmitting delay limitation as selective retransmission basic objects; using the transmitting end as a first communication node, determining importance levels of IP (Internet Protocol) packets in the DFS according to fault-tolerant importance measure, and supplying the selective retransmission of the IP packets by packet scheduling based on fault-tolerant importance; and on the basis of the importance levels of the IP packets, using a base station as a second communication node, and supplying the selective retransmission of a wireless link unit by a priority descending ARQ (Automatic Request for Repetition) mode. Under delay and bandwidth condition limitation, the invention organically combines priority transmission mechanisms among main communication nodes, realizes the importance division and the selective retransmission of wireless link unit levels, can strengthen the error control performance of online video streams and effectively increase the rebuilding quality of received videos.
Description
Technical field
The present invention relates to field of picture communication, particularly relate to a kind of Online Video stream selectivity repeating method of cross-node control.
Background technology
Along with the development of the Image Communication and the broadband cellular communication technology (as the B3G/4G cellular communication in the evolution), the honeycomb video stream of interconnection agreement Network Based (Internet Protocol is called for short " IP ") is used and is risen gradually.Be different from program request stream transmission based on the off-line video coding, the online application model of honeycomb video stream (being called for short " Online Video stream ") is under unique user acceptable start delay restriction, transmitting terminal is the single video of online coding on one side, the playback on one side receiving terminal is prescribed a time limit.This type of new application that develops into of advanced video coding and new generation broadband wireless mobile communication provides possibility, but its error control faces bigger technological challenge.For the two-way cellular communications networks that allows to retransmit, whole system can be divided into wireless access network and wired core net two large divisions, and radio link layer need further be divided into littler Radio Link unit with the IP bag.For Online Video stream, the stream primitive is the minimum transmission units that can independently retransmit between each communication node, and specifically, the IP bag is the stream primitive of wired core net, and the Radio Link unit is the stream primitive of wireless access network.Be accompanied by spatial-temporal prediction technology complicated day by day in the video coding, video flowing differs greatly for the sensitivity of various flows primitive loss, thereby causes video in various degree to degrade.In cellular communication system, the wireless access mesh portions is the bottleneck of video flowing reliable transmission, because it has the lower bandwidth and the error rate of Geng Gao than wired core net.Video flowing often causes the serious decline of reconstruction video quality by the normal Radio Link unit loss that burst occurs behind the wireless access network.Therefore, the error control of Online Video stream is the key technical problem that must solve.As everyone knows, repeat requests (Automatic Repeat Request is called for short " ARQ ") and forward error correction (Forward Error Correction is called for short " FEC ") are two kinds of basic error control methods automatically.In cellular communication system, FEC is suitable for the error control of physical layer, and at present common have RS sign indicating number, LDPC sign indicating number, a Turbo code.After physical layer FEC provided basic error control ability, the ARQ mode can strengthen the error control performance of Online Video stream neatly in data link layer or network layer.
For the Online Video stream that allows certain time-delay, the selectivity retransmission technique is a kind of effective measures that improve its error control performance, it need according to the fault-tolerant difference of importance of each stream primitive, effectively divide limited transfer resource to be equipped with the global optimization that obtains reconstruction quality.Video flowing need be introduced image sets (Group Of Pictures is called for short " GOP ") structure usually so that allow random access, error control and multimedia synchronous.In GOP level classifying importance, I frame, P frame and B frame have the transmission priority that successively decreases usually, and further, scalable video encoder can produce more multi-layered time transmission priority according to projected relationship.But existing priority division methods is not considered transmission distortion, can't realize the classifying importance of Radio Link unit.Therefore, according to the correlation theory that the Shannon message source and channel separates, the key of carrying out the selectivity re-transmission then is to estimate GOP level transmission distortion effectively, to assess the fault-tolerant importance of current stream primitive in a basic object.The GOP level transmission distortion of a stream primitive can adopt the analysis-synthetic technology of encoding and decoding again to carry out calculated off-line, but because limiting factors such as amount of calculation and real-time, this analysis-synthetic technology is difficult to be applied to Online Video stream.Existing selectivity retransmits only can provide very limited fault-tolerant importance type, how to estimate the GOP level transmission distortion of each stream primitive online, divide and the selectivity re-transmission with the importance that realizes Radio Link unit precision, do not see relevant technical literature at present.As typical A RQ mechanism, mix automatic request retransmission (Hybrid-ARQ, abbreviation " HARQ ") technology has been used for 3G high-speed packet access (High-Speed Packet Access, abbreviation " HSPA ") network is to the error control of Radio Link unit, but the data link layer mechanism of HARQ is not considered the otherness of video content, can't distinguish the fault-tolerant importance of different radio link unit.
Find patent " wireless video streaming of use single layer coding and prioritized streaming " (publication number: CN101073268 through after the retrieval to similar prior art; Classification number: H04N7) propose to propose simultaneously the level of some or all frame of video according to them optionally to be sent based on bandwidth constraints for each encoded video frame distributes a priority level.This patent has only provided a kind of priority of application layer and has divided and transmission method, the selectivity that can't be applied to network layer or data link layer retransmits, described priority level only depends on a number of frames among the GOP, can't be applied to less stream primitives such as Radio Link unit; This patent has only been considered the decoding dependence between the frame of front and back in the GOP, does not consider the transmission distortion of reconstruction video, and the non-stationary property of video content causes said method can not accurately reflect the actual priority of coded frame.
Studies show that in recent years, the isolation that each interlayer of communication protocol stack has been broken in layer (Cross-layer) design to a certain extent of striding towards wireless video, the parameter of each layer, state information etc. are carried out extraction and analysis and information sharing, be optimized judgement and strategy adjustment in conjunction with video content, the optimization that realizes resource disposes the self adaptation that reaches channel fluctuation, thereby reaches overall best performance.The existing layer retransmission technique of striding still is confined to carry out between two nodes of communication network, do not consider the combined optimization mechanism between more communication nodes, is not difficult to adapt to the video content of non-stationary in the single video flowing and the channel conditions of burst error.In the cellular network of multiple users share, the base station often needs a plurality of video flowings are handled simultaneously, therefore only support the importance of low complex degree to divide and the selectivity re-transmission for single video flowing, the base station wishes that transmitting terminal can provide certain collaborative process mechanism to assist it to reduce amount of calculation.Therefore in cellular communication system, need provide a kind of priority repeating method, to improve the reconstruction quality of receiver, video at Online Video stream in conjunction with the information source and the characteristic of channel.
Summary of the invention
Technical problem to be solved by this invention provides a kind of Online Video stream selectivity repeating method of cross-node control, can strengthen the error control performance of live video stream, thereby effectively improve the reconstruction quality of receiver, video.
The technical solution adopted for the present invention to solve the technical problems is: a kind of Online Video stream selectivity repeating method of cross-node control is provided, may further comprise the steps:
(1) according to starting and the transmission delay restriction, transmitting terminal is divided into the basic object that time-delay constraint frame set (Delay-sensitive Frame Set is called for short " DFS ") one by one retransmits as selectivity with the single video flowing of online coding;
(2) transmitting terminal is as first communication node, and the packet scheduling of at first carrying out based on fault-tolerant importance to retransmit for DFS provides the selectivity of IP bag;
(3) based on the importance rate of IP bag, the base station is as the second communication node, and the automatic repeat requests mode that execution priority is successively decreased to retransmit for DFS provides the selectivity of Radio Link unit.
Described DFS selects to retransmit the frame of video segmentation that marks off for single video flowing, wherein, includes the frame of video of equal amount in each time-delay constraint frame set; At transmitting terminal, each frame of video is encoded into P IP bag, P 〉=1; In the base station, each IP bag further is packaged into K the Radio Link unit with regular length, K 〉=1.
The packet scheduling based on fault-tolerant importance in the described step (2) may further comprise the steps:
Obtain and store side information; Described side information comprises packing manner, mistake compensation way, DFS length and instant coding result;
The IP bag that will come from same DFS assembles a coding formation Q by coded sequence;
According to fault-tolerant importance estimate determine each IP wrap under relative importance among the coding formation Q, and give each IP bag a unique importance rate;
According to the order that importance rate successively decreases all the IP bags among the coding formation Q are resequenced, obtain a new queue and rearrangement Q ' thus;
The real time transmission protocol serial that will increase progressively is distributed to IP bag in the queue and rearrangement successively by the importance rate order of successively decreasing, and the order that increases progressively by real time transmission protocol serial in same time slot scheduling sends queue and rearrangement Q '.
Base station in the described step (3) comprises network layer agency and link layer agency; May further comprise the steps in the automatic repeat requests mode that described priority is successively decreased:
Receive queue and rearrangement Q ' by described network layer agency, and according to the order after the rearrangement of real time transmission protocol serial recovery IP bag, when queue and rearrangement Q ' finishes receiving or dispatch the time limit when reaching, the described link layer agency of described network layer agent advertisement begins queue and rearrangement Q ' is carried out the automatic repeat requests transmission that priority is successively decreased;
Link layer agency further is divided into the Radio Link unit with different priorities according to two step priority decision rules with queue and rearrangement Q ', and the order of successively decreasing according to priority is successively to the transmission of Radio Link unit trial property;
Whether the link layer agency will be sent according to the number of transmissions upper limit decision of this Radio Link unit before attempting sending a Radio Link unit.
The computational methods that described fault-tolerant importance is estimated are as follows: if can calculate GOP level transmission distortion according to side information, the importance rate of the IP bag that GOP level transmission distortion is big more so is high more; Otherwise according to default order, the IP of I frame type wraps, the IP of P frame type wraps and the respectively corresponding high, medium and low three kinds of importance rates of IP bag of B frame type, and the IP that elder generation encodes in same type bag has higher importance; Described GOP level transmission distortion be the transmission distortion of the coded frame that causes of IP bag loss and not after the transmission distortion addition of coded frame divided by this IP packet length.
The transmission distortion of described coded frame is the Pixel-level transmission distortion sum of all pixels in coded frame in the IP bag; Described Pixel-level transmission distortion is the product of pixel compensation mistake and error diffusion counting; Described pixel compensation mistake is the absolute error that transmitting terminal is rebuild the reconstruction pixel value of pixel value and former frame same position; Described pixel error diffusion counting be in coded frame directly or indirectly with reference to the total pixel number of this pixel, obtain by the pixel count that in coded frame, adds up iteratively on this pixel place prediction link.
The transmission distortion of described not coded frame is that IP bag is lost in not in the coded frame error diffusion and estimates, obtains by piecewise linearity attenuation model or inter macroblocks ratio attenuation model; In described piecewise linearity attenuation model, frame level transmission distortion from IP bag place frame to subsequent motion compensation disassociation frame presents the attenuation characteristic of linear decrease frame by frame in time, its decrement step size by to after the IP bag the frame level transmission distortion of coded frame carry out linear fit and obtain; In described inter macroblocks ratio attenuation model, frame level transmission distortion from IP bag place frame to subsequent motion compensation disassociation frame presents the attenuation characteristic that ratio is successively decreased in time, and the inter macroblocks ratio of coded frame obtains its scale factor that successively decreases by adding up recently; Described motion compensation disassociation frame is the frame of video that has direct or indirect motion compensated prediction relation with this IP bag place frame.
Described two step priority decision rules are used to the unique priority of each Radio Link unit distribution among the DFS, comprise following two rules: rule one, and the Radio Link unit that belongs to higher importance rate IP bag has higher priority; Rule two, the Radio Link unit in the same IP bag has the priority of successively decreasing by coded sequence.
The maximum that the described the number of transmissions upper limit is this Radio Link unit allows the number of transmissions; If the actual transmissions number of times of this Radio Link unit less than the described the number of transmissions upper limit, then sends this Radio Link unit; Otherwise, stop to send the Radio Link unit of current DFS, then be ready for sending the Radio Link unit of next DFS.
Beneficial effect
Owing to adopted above-mentioned technical scheme, the present invention compared with prior art, have following advantage and good effect: in the method, the main communication node of Online Video stream can be shared out the work and helped one another, stream primitive for the different communication level provides the selectivity retransmission mechanism jointly, has realized other priority transmission of Radio Link cell level.In addition, fault-tolerant importance is estimated and be can be used for online assessment different IP and wrap in fault-tolerant importance under the various time-delays constraint, can adapt to the video content of non-stationary.Than existing selectivity transmission plan; the present invention can transfer to the transmission distortion of single video flowing unessential relatively part among each DFS; thereby protected relative part and parcel among the same DFS, under time-delay and bandwidth constraints, improved the whole body reconstruction quality of Online Video stream effectively.
Description of drawings
Fig. 1 is the Online Video streaming system framework and the main communication node schematic diagram thereof of the inventive method;
Fig. 2 is a DFS structural representation of the present invention;
Fig. 3 is the schematic functional block of the packet scheduling based on fault-tolerant importance of the present invention;
Fig. 4 is the schematic diagram of the GOP level transmission distortion that adopted of importance classifying module of the present invention;
Fig. 5 is the workflow diagram of the ARQ that successively decreases of medium priority of the present invention;
Fig. 6 is the application schematic diagram of the inventive method.
Embodiment
Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiment only to be used to the present invention is described and be not used in and limit the scope of the invention.Should be understood that in addition those skilled in the art can make various changes or modifications the present invention after the content of having read the present invention's instruction, these equivalent form of values fall within the application's appended claims institute restricted portion equally.
The present invention relates to a kind of Online Video stream selectivity repeating method of cross-node control, can strengthen the error control performance of Online Video stream, may further comprise the steps: the single video flowing of online coding at first enters transmitting terminal, the coded frame sequence limits the DFS that is divided into one by one according to startup and transmission delay, as the basic object of selectivity re-transmission; Transmitting terminal estimates to determine the importance rate of different IP bag among the DFS according to fault-tolerant importance, and the selectivity re-transmission of IP bag is provided by the packet scheduling based on fault-tolerant importance; Further, the priority of different radio link unit among the DFS is determined according to two step priority decision rules in the base station, and the selectivity re-transmission of Radio Link unit is provided by the ARQ mode that priority is successively decreased.
Fig. 1 is Online Video streaming system framework and the main communication node thereof that is suitable for the inventive method, and the selectivity of cross-node retransmits and finished jointly by transmitting terminal and these two communication nodes of base station.H.264/AVC Online Video stream come from or MPEG-4 etc. adopts the encoder of motion compensated prediction, the packet scheduling that transmitting terminal is carried out based on fault-tolerant importance provides the selectivity of IP bag to retransmit, on this basis, the execution priority ARQ that successively decreases in base station provides the selectivity of Radio Link unit to retransmit.
Fig. 2 is the structural representation of DFS among Fig. 1, and DFS is for Online Video stream selectivity retransmits a frame of video segmentation that marks off, with the basic object that retransmits as the cross-node selectivity.For simplicity of exposition, the specific implementation of video flowing GOP does not all comprise the B frame type among Fig. 2, Fig. 4 and Fig. 6.Along with the increase of cellular network bandwidth, the about hundreds of millisecond of the whole transmission delay of single video flowing and even lower, along with the lifting of processor performance, the actual coding time of a frame of video or decode time are lower than 33 milliseconds all the more.Single video stream application allows about several seconds startup time-delay usually, and the frame of video of online coding is at interval greatly between 33 milliseconds to 100 milliseconds, therefore each DFS can contain some frame of video according to the time-delay constraint, and the time-delay constraint should be less than starting the poor of time-delay and transmission delay.Under given time-delay constraint, each DFS contains the frame of video of equal amount.At transmitting terminal, each frame is encoded into some IP bags; In the base station, each IP bag further is packaged into some Radio Links unit.DFS is two different notions with GOP.For convenience of description, the length " N of DFS
D" (N
D〉=1) represents the number of frames of each DFS, N among Fig. 2
DEqual 3; " N
G" represent the number of frames of each GOP, its representative value has 12,24, and 15 and 30.N
DValue can be greater than N
G, such DFS just comprises the frame of a plurality of GOP.Frame of video and image sheet are numbered according to coded sequence, the frame number that " n " (n 〉=1) expression increases progressively according to coded sequence, and frame of video n is split into " M
n" (M
n〉=1) individual image sheet, " P
N, m" expression n frame of video " m " (1≤m≤M
n) the pairing IP bag of individual image sheet, " S
N, m" expression P
N, mThe byte number of packet length.In wireless access network, P
N, mBe split into the Radio Link unit that several have regular length, " K
N, m" expression P
N, mThe number of contained Radio Link unit, its value equals the ratio of IP packet length and Radio Link element length.
Fig. 3 is based on the schematic functional block of the packet scheduling of fault-tolerant importance among Fig. 1.For Online Video stream, encoder and transmitting terminal are the communication nodes of two loose couplings.Transmitting terminal utilizes the side information extraction module to obtain and store side information.Wherein, side information comprises packing manner, mistake compensation way, DFS length and instant coding result, and these information will offer the importance classifying module.Particularly, packing manner is according to the subpackage rule of RFC 3894, the mode that frame duplicated before the mistake compensation way adopted, DFS length N
DDetermine according to the time-delay constraint that current application allowed.The IP bag of encoder output at first is input to frame set scheduler module, and this module is divided into one by one DFS according to coded sequence with a certain video flowing, will assemble the formation of a coding according to coded sequence from the IP bag of same DFS
(p≤n＜" p+N
D") ".In order to determine the relative importance of each IP bag among the current DFS, the importance classifying module need be calculated P
N, mGOP level transmission distortion." GTD (n, m) " expression IP bag P
N, mLose the GOP level transmission distortion that causes, i.e. transmission distortion sum separately from the n frame to this GOP last frame.
Fig. 4 is the schematic diagram of the GOP level transmission distortion that the importance classifying module is adopted among Fig. 3, and the loss of an IP bag will cause error of transmission to be spread in reconstruction video, till the last frame of this IP bag place GOP.As shown in Figure 4, the mistake that the black rectangle zone causes in the frame 2 can spread on time of reconstruction video and direction in space, till an I frame occurring and eliminating this influence.For convenience, " N
c" last frame number of coded frame of the current DFS of expression, " N
g" expression P
N, mThe frame number of the last frame of place GOP, P
N, mPlace GOP may be coded frame entirely.Frame N
c(comprise frame N before
c) transmission distortion need analyze the instant coding result that the side information extraction module provides, frame N
cTransmission distortion afterwards then needs to estimate to obtain by certain forecast model." GTD
e(n, m) " (n≤N
c) expression IP bag P
N, mThe transmission distortion of the coded frame that loss causes, " GTD
f(n, m) " (n＞N
c) expression IP bag P
N, mThe transmission distortion of the not coded frame that loss causes." i
N, m" expression P
N, mIn a pixel,
Expression i
N, mAt the reconstruction pixel value of transmitting terminal, can from the coding link of encoder, obtain.Absolute error average (being called for short " MAD ") is a measurement criterion commonly used in the video coding, and its value is more little, and the expression performance is good more.If N
cLess than N
g, show that (n m) comprises the transmission distortion of coded frame and the transmission distortion of coded frame not to GTD; Otherwise (n m) only comprises the transmission distortion of coded frame to GTD.Particularly, the importance classifying module is calculated GTD (n, value m) according to formula (1).
Further, GTD
e(n m) is P
N, mIn the Pixel-level transmission distortion sum of all pixels in coded frame.The Pixel-level transmission distortion is the product of pixel compensation mistake and error diffusion counting: the pixel compensation mistake is the absolute error that transmitting terminal is rebuild the reconstruction pixel value of pixel value and former frame same position; The error diffusion of a certain pixel counting is directly or indirectly with reference to the total pixel number (comprising this pixel itself) of this pixel in coded frame, promptly by following the tracks of final this pixel motion vector information of pointing in the coded frame, totally point to the sum of all pixels on all prediction links of this pixel iteratively.
Remarked pixel i
N, mError of compensation, " SRC
k(i
N, m) " expression i
N, mThe error diffusion counting from frame n to frame k that loss is caused can be accumulated i by iteration
N, mNumber of pixels on the motion compensated prediction path, place obtains, and the both belongs to side information.Because the Pixel-level transmission distortion can add up in an IP bag, the importance classifying module is calculated GTD according to formula (2)
e(n, value m).
GTD
f(n m) is P
N, mThe error diffusion distortion that is lost in the coded frame is not estimated, can't obtain by the method for pursuit movement vector.The invention provides two kinds of models and estimate GTD respectively
f(n, m) value, i.e. piecewise linearity attenuation model or inter macroblocks ratio attenuation model, wherein the piecewise linearity attenuation model is estimated often more accurately, but complexity is higher.
In the piecewise linearity attenuation model, from P
N, mThe place frame presents the attenuation characteristic of linear decrease frame by frame in time to the transmission distortion of subsequent motion compensation disassociation frame, " λ (n, m) " expression P
N, mLoss causes the not decrement step size of the frame level transmission distortion of coded frame, and its value through type (3) is counted the error diffusion of coded frame and carried out linear fit and obtain.
Based on above-mentioned decrement step size, the importance classifying module can be estimated from frame " N by through type (4)
c+ 1 " frame N in the same GOP
gThe transmission distortion sum:
In the following formula, consider " n=N
c" special circumstances, " λ (l, m) " (expression of l＜n) is from P
N, mNearest IP bag P
L, mPairing decrement step size.
In inter macroblocks ratio attenuation model, P
N, mThe frame level transmission distortion that loss causes is at frame N
cPresent the attenuation characteristic that ratio is frame by frame successively decreased in the follow-up motion compensation disassociation frame, its successively decrease scale factor by statistics recently the inter macroblocks ratio of coded frame obtain.In adopting the Online Video stream of motion compensated predictive coding, all macro blocks can be classified as two big classes: based on the inter macroblocks of interframe movement estimation and the intra-frame macro block of no interframe movement estimation.Inter macroblocks ratio among " r (n) " expression frame n, the i.e. ratio of inter macroblocks quantity and every frame macro block sum.n
1And n
2(n
2＜n
1＜N
c) represent from frame N
cTwo nearest P frames of having encoded.Based on above-mentioned inter macroblocks ratio, the importance classifying module also can be estimated from frame " N by through type (5)
c+ 1 " frame N in the same GOP
gThe transmission distortion sum:
In the following formula, " β
1" and " β
2" (β
1+ β
2=1) be two weighting parameters, initial value is respectively 0.8 and 0.2.After finishing this frame coding, β
1And β
2Can upgrade by the least square fitting mode according to the inter macroblocks ratio of reality.
In same DFS, the IP bag that GOP level transmission distortion is big more has higher importance rate.(n m) determines IP bag P to the importance classifying module according to GTD
N, mRelative importance in affiliated DFS, and give P
N, mUnique importance rate " c (n, m) ".Then, the order that the bag reordering module is successively decreased according to importance rate sorts all the IP bags among the coding formation Q, forms a new queue and rearrangement thus
(q≤n＜" q+N
D"), wherein the IP that importance rate is high wraps in the front of queue and rearrangement Q ', P
Q, 1Has the highest importance rate.Forming Q ' afterwards, the bag reordering module is immediately with P
Q, 1With
Sequence number send to the base station in advance as signaling information.Then, the RTP that string and modular converter will increase progressively (being called for short " RTP ") sequence number is distributed to all the IP bags among the Q ' successively, and the order that increases progressively by the RTP sequence number in same time slot scheduling sends these IP bags.After transmission finishes, queue and rearrangement Q ' will be buffered a period of time to be used for the selectivity re-transmission of IP bag at transmitting terminal.Based on the DFS structure, 16 RTP sequence number can provide 65536 importance rates, and this is far above 64 importance rates of maximum that H.264/SVC provide at present.
Need to prove, in the time can't calculating GOP level transmission distortion according to side information, can be according to default order to unique importance rate of IP bag, the IP that is about to the I frame type wraps, the IP of P frame type wraps and the respectively corresponding high, medium and low three kinds of importance rates of IP bag of B frame type, and the IP that elder generation encodes in same frame type bag has higher importance rate.
After transmitting terminal provides the selectivity re-transmission of IP bag precision, the base station will further provide the selectivity of Radio Link unit precision to retransmit.The network layer of base station agency is used to receive the IP bag of queue and rearrangement Q ', and the order when recovering IP and wrap in string and modular converter according to the RTP sequence number.Suppose that queue and rearrangement Q ' contains C
Q 'Individual IP bag wherein only has the IP bag of A% to be allowed to retransmit.In the scheduling time limit of this Q ', the network layer agency will select the wherein higher [C of importance rate
Q '* A%] individual IP bag request retransmission, send corresponding negative response (Negative Acknowledgement is called for short " NACK ") bag and give transmitting terminal, the NACK bag adopts RFC 4585 standards.Be limited to when queue and rearrangement Q ' finishes receiving or dispatches, the network layer agency notifies the link layer agency to begin the ARQ transmission mechanism that current DFS execution priority is successively decreased immediately.
The link layer agency of base station wraps P based on the data link layer technology of wireless access network with IP
N, mFurther be divided into K
N, mIndividual Radio Link unit with regular length, current like this DFS has
Individual Radio Link unit, the link layer agency is this L according to two step priority decision rules
Q 'Individual Radio Link unit distributes unique priority.Wherein, two step priority decision rules are used to the unique priority of each Radio Link unit distribution among the DFS, comprise following two rules: rule one, and the Radio Link unit that belongs to higher importance rate IP bag has higher priority; Rule two, the Radio Link unit in the same IP bag has the priority of successively decreasing by coded sequence.According to coded sequence, " l
N, m, k" expression IP bag P
N, m" k " (1≤k≤K
N, m) individual Radio Link unit.The Radio Link unit l that from a DFS, has limit priority
Q, 1,1Beginning, the order that link layer agency successively decreases according to priority are successively to the transmission of the Radio Link unit trial property of current DFS.
Fig. 5 is the workflow diagram of the ARQ that successively decreases of Fig. 1 medium priority.In wireless access network, about about a few tens of milliseconds, if do not receive the ARQ information of current wireless link unit behind round trip delay time, the link layer agency will send the next Radio Link unit of video flowing to the round trip delay time of data link frame immediately greatly.A certain Radio Link unit has the transmission opportunity of certain number of times, and link layer is acted on behalf of the actual transmissions number of times of this Radio Link unit that need add up." A
N, m, k" (A
N, m, k〉=0) expression Radio Link unit l
N, m, kThe actual transmissions number of times, initial value is 0, every transmission l
N, m, kOnce, A
N, m, kAdd up 1." Φ
N, m, k" (Φ
N, m, k〉=0) expression Radio Link unit l
N, m, kThe number of transmissions upper limit, this value by link layer agency at the first l that sends
N, m, kBefore calculate according to formula (6).
In the following formula, real number " η " (η 〉=0) is represented the maximum average transmission number of times that each Radio Link unit is allowed, and it is illustrated in link layer transfer resource total in the wireless access network.Symbol
Expression is not more than the maximum integer of x.In the ARQ mechanism that priority is successively decreased, if current Φ
N, m, kEqual 0, the link layer agency stops to send the Radio Link unit of current DFS, just abandons all Radio Link unit that do not pass among the current DFS, then is ready for sending the Radio Link unit of next DFS; If current Φ
N, m, kGreater than 0, the link layer agency will attempt sending l
N, m, k, if the physical layer error that can't correct causes l
N, m, kLoss, receiving terminal will feed back the ARQ information of this Radio Link unit immediately, and the link layer agency can be to l
N, m, kCarry out " Φ at most
N, m, k-1" inferior re-transmission.If Radio Link unit l
N, m, kBe successfully received, the link layer agency will calculate next Radio Link unit " l according to formula (6)
N, m, k+1" the number of transmissions upper limit " Φ
N, m, k+1".If the L of current DFS
Q 'Individual Radio Link unit all is successfully received, and the link layer agency of base station will carry out the transmission of the transmission of next DFS and prepare.If the maximum transmission times that current DFS allows has been used up in the Radio Link unit of higher priority, the link layer agency has to abandon the Radio Link unit of some lower priorities.
At last, the Application Example in conjunction with Fig. 6 introduces a simplification of the present invention is used to further specify thought of the present invention.Each DFS contains 3 frame of video (N successively
D=3), each frame of video is encoded into 2 IP bags with close length, and each IP bag further is packaged into 2 Radio Link unit.Finish the coding of the 3rd frame at encoder after, 6 IP bags among first DFS will be formed a coding formation Q={P
1,1P
1,2P
2,1P
2,2P
3,1P
3,2.By GOP level transmission distortion more separately, each the IP bag among first DFS will be assigned with a unique importance rate, i.e. { c (1,1)=6; C (1,2)=5; C (2,1)=2; C (2,2)=1; C (3,1)=3; C (3,2)=4}.(n, the order of m) packet-by-packet successively decreasing, bag reordering module rearrange this 6 IP bags, form a new queue and rearrangement Q '={ P thus according to GTD
1,1P
1,2P
3,2P
3,1P
2,1P
2,2.The network layer agency who supposes the base station intactly receives this queue and rearrangement Q ', and η equals 1.Although the physical layer error period appears at Radio Link unit l
1,1,2, l
1,2,2, l
3,2,1, l
3,2,2Transmission course in, the ARQ mechanism that priority is successively decreased will retransmit these Radio Link unit as possible according to its number of transmissions upper limit.Radio Link unit at higher priority (is l
1,1,2, l
1,2,2, l
3,2,1, l
3,2,2) be successfully received after, the Radio Link unit that link layer agency initiatively abandons lower priority (is l
2,1,1, l
2,1,2, l
2,2,1, l
2,2,2) with the limited total transfer resource of balance.Like this, unessential relatively part among the DFS is finally transferred in error of transmission, has realized the selectivity of single video flowing is retransmitted under non-stationary video content and Burst Channel error situation.
Claims (9)
1. the Online Video stream selectivity repeating method of a cross-node control is characterized in that, may further comprise the steps:
(1) according to starting and the transmission delay restriction, transmitting terminal is divided into the basic object that time-delay constraint frame set one by one retransmits as selectivity with the single video flowing of online coding;
(2) transmitting terminal is as first communication node, and the packet scheduling of at first carrying out based on fault-tolerant importance to provide the selectivity of IP bag to retransmit for the set of time-delay constraint frame;
(3) based on the importance rate of IP bag, the base station is as the second communication node, and the automatic repeat requests mode that execution priority is successively decreased to provide the selectivity of Radio Link unit to retransmit for the set of time-delay constraint frame.
2. the Online Video stream selectivity repeating method of cross-node control according to claim 1, it is characterized in that, the set of described time-delay constraint frame is to select to retransmit the frame of video segmentation that marks off for single video flowing, wherein, the frame of video that includes equal amount in each time-delay constraint frame set; At transmitting terminal, each frame of video is encoded into P IP bag, P 〉=1; In the base station, each IP bag further is packaged into K the Radio Link unit with regular length, K 〉=1.
3. the Online Video stream selectivity repeating method of cross-node control according to claim 1 is characterized in that the packet scheduling based on fault-tolerant importance in the described step (2) may further comprise the steps:
Obtain and store side information; Described side information comprises packing manner, mistake compensation way and time-delay constraint frame set length and instant coding result;
The IP bag that will come from the set of same time-delay constraint frame assembles a coding formation Q by coded sequence;
Estimate to determine that according to fault-tolerant importance each IP wraps in the relative importance among the affiliated coding formation Q, and give each IP bag a unique importance rate;
According to the order that importance rate successively decreases all the IP bags among the coding formation Q are resequenced, obtain a new queue and rearrangement Q ' thus;
The real time transmission protocol serial that will increase progressively is distributed to IP bag in the queue and rearrangement successively by the importance rate order of successively decreasing, and the order that increases progressively by real time transmission protocol serial in same time slot scheduling sends queue and rearrangement Q '.
4. the Online Video stream selectivity repeating method of cross-node control according to claim 1 is characterized in that the base station in the described step (3) comprises network layer agency and link layer agency; May further comprise the steps in the automatic repeat requests mode that described priority is successively decreased:
Receive queue and rearrangement Q ' by described network layer agency, and according to the order after the rearrangement of real time transmission protocol serial recovery IP bag, when queue and rearrangement Q ' finishes receiving or dispatch the time limit when reaching, the described link layer agency of described network layer agent advertisement begins queue and rearrangement Q ' is carried out the automatic repeat requests transmission that priority is successively decreased;
Link layer agency further is divided into the Radio Link unit with different priorities according to two step priority decision rules with queue and rearrangement Q ', and the order of successively decreasing according to priority is successively to the transmission of Radio Link unit trial property;
Whether the link layer agency will be sent according to the number of transmissions upper limit decision of this Radio Link unit before attempting sending a Radio Link unit.
5. the Online Video stream selectivity repeating method of cross-node control according to claim 3 is characterized in that the computational methods that described fault-tolerant importance is estimated are as follows:
If can computed image group level transmission distortion according to side information, the importance rate of the IP bag that image sets level transmission distortion is big more so be high more; Otherwise according to default order, the IP of I frame type wraps, the IP of P frame type wraps and the respectively corresponding high, medium and low three kinds of importance rates of IP bag of B frame type, and the IP that elder generation encodes in same type bag has higher importance; Described image sets level transmission distortion be the transmission distortion of the coded frame that causes of IP bag loss and not after the transmission distortion addition of coded frame divided by this IP packet length.
6. the Online Video stream selectivity repeating method of cross-node according to claim 5 control is characterized in that, the transmission distortion of described coded frame is the Pixel-level transmission distortion sum of all pixels in coded frame in the IP bag; Described Pixel-level transmission distortion is the product of pixel compensation mistake and error diffusion counting; Described pixel compensation mistake is the absolute error that transmitting terminal is rebuild the reconstruction pixel value of pixel value and former frame same position; Described pixel error diffusion counting be in coded frame directly or indirectly with reference to the total pixel number of this pixel, obtain by the pixel count that in coded frame, adds up iteratively on this pixel place prediction link.
7. the Online Video stream selectivity repeating method of cross-node control according to claim 5, it is characterized in that, the transmission distortion of described not coded frame is that IP bag is lost in not in the coded frame error diffusion and estimates, obtains by piecewise linearity attenuation model or inter macroblocks ratio attenuation model; In described piecewise linearity attenuation model, frame level transmission distortion from IP bag place frame to subsequent motion compensation disassociation frame presents the attenuation characteristic of linear decrease frame by frame in time, its decrement step size by to after the IP bag the frame level transmission distortion of coded frame carry out linear fit and obtain; In described inter macroblocks ratio attenuation model, frame level transmission distortion from IP bag place frame to subsequent motion compensation disassociation frame presents the attenuation characteristic that ratio is successively decreased in time, and the inter macroblocks ratio of coded frame obtains its scale factor that successively decreases by adding up recently; Described motion compensation disassociation frame is the frame of video that has direct or indirect motion compensated prediction relation with this IP bag place frame.
8. the Online Video stream selectivity repeating method of cross-node control according to claim 4, it is characterized in that, described two step priority decision rules are used to the unique priority of each Radio Link unit distribution in the set of time-delay constraint frame, comprise following two rules: rule one, the Radio Link unit that belongs to higher importance rate IP bag has higher priority; Rule two, the Radio Link unit in the same IP bag has the priority of successively decreasing by coded sequence.
9. the Online Video stream selectivity repeating method of cross-node control according to claim 4 is characterized in that the maximum that the described the number of transmissions upper limit is this Radio Link unit allows the number of transmissions; If the actual transmissions number of times of this Radio Link unit less than the described the number of transmissions upper limit, then sends this Radio Link unit; Otherwise, stop to send the Radio Link unit of current delay constraint frame set, then be ready for sending the Radio Link unit of next time-delay constraint frame set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010284648 CN101938341B (en) | 2010-09-17 | 2010-09-17 | Cross-node controlled online video stream selective retransmission method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010284648 CN101938341B (en) | 2010-09-17 | 2010-09-17 | Cross-node controlled online video stream selective retransmission method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101938341A true CN101938341A (en) | 2011-01-05 |
| CN101938341B CN101938341B (en) | 2012-12-05 |
Family
ID=43391493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010284648 Expired - Fee Related CN101938341B (en) | 2010-09-17 | 2010-09-17 | Cross-node controlled online video stream selective retransmission method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101938341B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103621085A (en) * | 2011-06-30 | 2014-03-05 | 微软公司 | Reducing latency in video encoding and decoding |
| WO2014139427A1 (en) * | 2013-03-12 | 2014-09-18 | Huawei Technologies Co., Ltd. | System and method for multi-layer protocol selection |
| CN105491402A (en) * | 2015-12-08 | 2016-04-13 | 中国电子科技集团公司第二十八研究所 | Self-adaptive scalable video transmission method based on partial cooperative communication protocol |
| CN105744357A (en) * | 2016-02-29 | 2016-07-06 | 哈尔滨超凡视觉科技有限公司 | Method for reducing network video bandwidth occupation based on online resolution improvement |
| CN106028033A (en) * | 2016-05-10 | 2016-10-12 | 天津大学 | Self-adaptive cross-layer forward error correction method applicable to wireless video transmission |
| WO2019084970A1 (en) * | 2017-11-06 | 2019-05-09 | 华为技术有限公司 | Packet forwarding method, forwarding device, and network device |
| WO2020207180A1 (en) * | 2019-04-12 | 2020-10-15 | 展讯通信（上海）有限公司 | Data frame retransmission method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101073268A (en) * | 2004-12-10 | 2007-11-14 | 皇家飞利浦电子股份有限公司 | Wireless video streaming using single layer coding and prioritized streaming |
| CN101179362A (en) * | 2006-11-07 | 2008-05-14 | 中兴通讯股份有限公司 | Automatic retransmission request mechanism suitable for mobile stream media application |
-
2010
- 2010-09-17 CN CN 201010284648 patent/CN101938341B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101073268A (en) * | 2004-12-10 | 2007-11-14 | 皇家飞利浦电子股份有限公司 | Wireless video streaming using single layer coding and prioritized streaming |
| CN101179362A (en) * | 2006-11-07 | 2008-05-14 | 中兴通讯股份有限公司 | Automatic retransmission request mechanism suitable for mobile stream media application |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9729898B2 (en) | 2011-06-30 | 2017-08-08 | Mircosoft Technology Licensing, LLC | Reducing latency in video encoding and decoding |
| CN103621085A (en) * | 2011-06-30 | 2014-03-05 | 微软公司 | Reducing latency in video encoding and decoding |
| CN105721880A (en) * | 2011-06-30 | 2016-06-29 | 微软技术许可有限责任公司 | Reducing Latency In Video Encoding And Decoding |
| US9426495B2 (en) | 2011-06-30 | 2016-08-23 | Microsoft Technology Licensing, Llc | Reducing latency in video encoding and decoding |
| US9743114B2 (en) | 2011-06-30 | 2017-08-22 | Microsoft Technology Licensing, Llc | Reducing latency in video encoding and decoding |
| CN103621085B (en) * | 2011-06-30 | 2016-03-16 | 微软技术许可有限责任公司 | Reduce method and the computing system of the delay in video decode |
| CN105721880B (en) * | 2011-06-30 | 2019-04-05 | 微软技术许可有限责任公司 | For reducing the method and system of the delay in Video coding and decoding |
| US10003824B2 (en) | 2011-06-30 | 2018-06-19 | Microsoft Technology Licensing, Llc | Reducing latency in video encoding and decoding |
| WO2014139427A1 (en) * | 2013-03-12 | 2014-09-18 | Huawei Technologies Co., Ltd. | System and method for multi-layer protocol selection |
| CN105491402B (en) * | 2015-12-08 | 2018-04-10 | 中国电子科技集团公司第二十八研究所 | A kind of adaptive scalable video transmission method based on partial cooperative communication protocol |
| CN105491402A (en) * | 2015-12-08 | 2016-04-13 | 中国电子科技集团公司第二十八研究所 | Self-adaptive scalable video transmission method based on partial cooperative communication protocol |
| CN105744357A (en) * | 2016-02-29 | 2016-07-06 | 哈尔滨超凡视觉科技有限公司 | Method for reducing network video bandwidth occupation based on online resolution improvement |
| CN106028033B (en) * | 2016-05-10 | 2019-01-18 | 天津大学 | A kind of adaptive cross-layer forward error correction suitable for wireless video transmission |
| CN106028033A (en) * | 2016-05-10 | 2016-10-12 | 天津大学 | Self-adaptive cross-layer forward error correction method applicable to wireless video transmission |
| US11677671B2 (en) | 2017-11-06 | 2023-06-13 | Huawei Technologies Co., Ltd. | Time-sensitive networking (TSN) packet forwarding |
| WO2019084970A1 (en) * | 2017-11-06 | 2019-05-09 | 华为技术有限公司 | Packet forwarding method, forwarding device, and network device |
| US11245634B2 (en) | 2017-11-06 | 2022-02-08 | Huawei Technologies Co., Ltd. | Time-sensitive networking (TSN) packet forwarding |
| WO2020207180A1 (en) * | 2019-04-12 | 2020-10-15 | 展讯通信（上海）有限公司 | Data frame retransmission method and device |
| US11424870B2 (en) | 2019-04-12 | 2022-08-23 | Spreadtrum Communications (Shanghai) Co., Ltd. | Data frame retransmission method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101938341B (en) | 2012-12-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101938341B (en) | Cross-node controlled online video stream selective retransmission method | |
| Wu et al. | Bandwidth-efficient multipath transport protocol for quality-guaranteed real-time video over heterogeneous wireless networks | |
| AU2007313931B2 (en) | Dynamic modification of video properties | |
| Wu et al. | Improving multipath video transmission with raptor codes in heterogeneous wireless networks | |
| CN100539544C (en) | Medium streaming distribution system | |
| CN101636983B (en) | Reducing effects of packet loss in video transmissions | |
| Xu et al. | CMT-NC: improving the concurrent multipath transfer performance using network coding in wireless networks | |
| CN1859579B (en) | Apparatus and method for transmitting a multimedia data stream | |
| CN101304537B (en) | 3g-ip wireless video transmission system | |
| CN103840917A (en) | Multipath parallel transmission scheme based on network coding | |
| CN101826937B (en) | Link layer error control system suitable for next generation mobile Internet and method thereof | |
| CN101938770A (en) | Wireless network maximum retransmission times optimization method based on network channel state | |
| CN103051981A (en) | Adaptive frame size calculation method for multiple hop transmission of video | |
| CN109155707A (en) | The request data retransmission in multicast network | |
| CN102160340B (en) | Transmission rate control device and transmission rate control method | |
| Khalaf et al. | Improving video transmission over heterogeneous network by using ARQ and FEC error correction algorithm | |
| Lin et al. | An access point-based FEC mechanism for video transmission over wireless LANs | |
| CN103354615B (en) | Signal intensity based live video data transmission error control method | |
| Yao et al. | IPB-frame adaptive mapping mechanism for video transmission over IEEE 802.11 e WLANs | |
| CN103338412A (en) | Adaptive video coding systemcwireless adaptive modulation and coding | |
| Wu et al. | CoRET: a network coding based multicast retransmission scheme for mobile communication networks | |
| CN103339912A (en) | A video packet scheduling method for multimedia streaming | |
| Sakr et al. | Performance evaluation using smart: HARQ versus HARQ mechanisms beyond 5G networks | |
| CN103684693B (en) | Opportunistic network encoding implementation method for wireless network video perception | |
| CN105516109A (en) | Transmission scheduling method for ensuring video transmission service quality of wireless multi-hop network |
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: 20121205 Termination date: 20150917 |
|
| EXPY | Termination of patent right or utility model |