CN105007541A - Scalable video stream dynamic multi-rate multicast optimal transmission method - Google Patents

Abstract

The invention provides a scalable video stream dynamic multi-rate multicast optimal transmission method based on opportunistic routing and network coding. According to the method, the opportunistic routing mechanism of relay node forwarding is combined with a network coding technology, and meanwhile the time-varying dynamic characteristic of a wireless network, the competition relation of a wireless relay node and the code stream priority problem of a video coding layer are considered, so that the demand on dependence among scalable video coding layers is met in real time. A fully distributed code rate allocation method is adopted, so that joint optimization of dynamic code rate resource allocation, dynamic transmission scheduling and dynamical route selection of scalable video stream dynamic multi-rate multicast transmission is realized. Through adoption of the scalable video stream dynamic multi-rate multicast optimal transmission method, the overall throughput of the network is increased; the adaptability to the time-varying channel state is enhanced; and better video quality is provided for a receiving end.

Description

Telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method

Technical field

The present invention relates to a kind of method of data communication technology field, particularly, relate to a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method based on chance route and network code.

Background technology

Along with increasing rapidly and intelligent terminal day by day universal of Mobile data flow, radio video streaming media technology obtains in recent years and applies more and more widely.The analysis report of Cisco System Co. of the U.S. to global Mobile data shows that, by the end of the year 2014, mobile video data traffic has occupied more than 50% of Mobile data total flow, and the expection of this ratio will reach 75% in the end of the year 2019.But the transfer of data in wireless network is often easily subject to the impact of wireless channel time-varying characteristics and error of transmission, therefore how in the wireless network for transmission of video provides quality of service guarantee to remain a challenging problem.

As a kind of important means of video content distribution in network, the multi code Rate of Chinese character multicasting technology of scalable video (SVC, Scalable VideoCoding) Streaming Media can adapt to different users and watch the network condition of demand and isomery to improve the efficiency of transmission of network.The code stream of scalable video comprises a Primary layer and multiple enhancement layer, and the hierarchy of these flexible multidimensional provides multiple access point in reconstruction quality three dimensions of spatial resolution, time domain frame speed and video.When transmitting scalable video stream in the mode of multi code Rate of Chinese character multicast, each scalable video layer is transmitted by different IP multicast group, each recipient selects by its different disposal ability and different link capacities the multicast group adding some, thus obtains the sequence of video images of same content under different scale combination.

Through finding the retrieval of prior art, the people such as J.Zhao are at " IEEE Transactions on Multimedia, Oct.2006, pp.1021-1032, (institute of electrical and electronic engineers multimedia journal, in October, 2006, 1021-1032 page) " on delivered and be entitled as the article of " LION:Layered overlay multicast with network coding (LION: the layered coverage multicast using network code) ", this article uses the code check of video reception and the corresponding quality of reception as the utility index of terminal use, multi code Rate of Chinese character multicast for layered video stream constructs overall network utility maximization problems (NUM, Network Utility Maximization), when meeting extensible video stream interlayer dependency constraint and network capacity constraint, maximize the overall video quality of reception of all terminal uses for all telescopic video layers.It is to be noted, this article mainly based on static network characteristic parameter (such as topology, link capacity etc.) and time-independent hypothesis, and at one or more transmission path carrying out NUM Optimization Modeling and must pre-determining before solving from information source node to each terminal use.But, in the wireless video transmission network of reality, due to the time-varying characteristics of wireless channel, the characteristic parameter of wireless network often changes continually in time, and the optimal rate-allocation scheme of static network hypothesis that what therefore this article proposed be suitable for can not obtain optimum transmission of video performance.

Also find through retrieval, the people such as K.Zeng are at " IEEE Transactions on Wireless Communication, Dec.2008, pp.5118-5128, (institute of electrical and electronic engineers radio communication journal, in December, 2008, 5118-5128 page) " on delivered and be entitled as the article of " Capacity of opportunistic routing in multi-rate and multi-hop wirelessnetworks the capacity of chance route (in multi code Rate of Chinese character and the multi-hop wireless network) ", this article points out that chance method for routing can utilize the via node that the broadcast characteristic of wireless shared media and space diversity characteristic are data packet transmission dynamically to select next-hop node, compared to traditional static routing mechanism, significantly improve the end-to-end transmission throughput of wireless time-varying network.But the method that this article proposes only have studied from source node to the clean culture scene of some concrete terminal nodes in network, do not consider how to improve the occasion routing mechanism realization from source node to the multicast transmission optimization of multiple heterogeneous terminals user.

Summary of the invention

For defect of the prior art, the object of this invention is to provide a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method based on chance route and network code, the method combines chance routing mechanism and the network coding technique of via node forwarding, taken into account simultaneously wireless network time variation step response, the competitive relation of wireless relay nodes and the code stream issue of priority of video coding layer, to meet the dependent demand of scalable video interlayer in real time, a kind of complete distributed code rate allocation method is provided, finally realize the dynamic code rate Resourse Distribute of extensible video stream multi code Rate of Chinese character cast communication, dynamics route selection moves and the combined optimization of state transmitting and scheduling.

For realizing above object, the invention provides a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method based on chance route and network code, comprising the following steps:

The first step, at source node place, time-varying characteristics according to wireless network carry out segmentation to the time, within each time period, scalable video technology is used video flowing to be encoded to multiple scalable video layer, by ensureing the interlayer dependency constraint condition of scalable video to the Data Rate Distribution of different scalable video layer;

Second step, at via node place, when some concrete video datas wrap in and forward via the via node in network, use chance routing mechanism dynamically to select down hop forward node for it, and use network coding technique to improve the throughput of network further;

3rd step, in the wireless network, each wireless network node in wireless shared media is divided into several maximum while transmission node set, realize the overall network transmission time each maximum while transmission node set between optimal scheduling;

4th step, in conjunction with the interlayer dependent form condition of the source node place video coding layer that three steps above obtain, the chance routing mechanism that via node forwards and network coding technique, and the time variation step response and node-node transmission competitive relation of wireless network, use the dynamic network maximization of utility modeling method of many time periods to realize the optimization problem of the telescopic video flowable state multi code Rate of Chinese character cast communication based on chance route and network code;

When carrying out concrete Data Rate Distribution scheme for terminal node, adopt complete distributed code rate allocation method, finally realize the combined optimization of the dynamic code rate Resourse Distribute of extensible video stream multi code Rate of Chinese character cast communication, dynamics route selection and dynamic transmission scheduling.

Preferably, in the first step, described time slice is the Further Division to extensible video stream groupcast time, and in each time slice, the channel condition information of wireless channel is relatively stable, and the method passing through directly measurement and channel state prediction obtains.

Preferably, in the first step, the interlayer dependency constraint condition of described scalable video is that the decoding of the video layer that label is higher needs to depend on the lower video layer of label, and namely the multicast transmission order of video layer needs to carry out according to video layer label order from low to high.

Preferably, in second step, described chance routing mechanism is the broadcast characteristic and the space diversity characteristic that utilize wireless shared media, forwards in both candidate nodes set dynamically for packet selects actual down hop forward node in the down hop of via node.

Preferably, in second step, described network coding technique carried out arithmetic coding operation to packet before via node forwards, and the actual bandwidth consumption amount specifying on every bar link is the maximum of all destination nodes bandwidth consumed on the link; This condition is the constraints adopting network code on link, to realize the resource-sharing on the same link of different destination node.

Preferably, in the 3rd step, when while described, in transmission node set, all nodes carry out data packet transmission at the same time, its link be associated can not produce wireless competition and can be successfully completed the data packet transmission of next-hop node.

Preferably, in the 3rd step, described maximum while transmission node set be a transmission node set simultaneously, if increase any one node in this set, cause this set to become nonsimultaneous transmission node set.

Preferably, in 4th step, described combined optimization problem is: the video total quality received in the complete period of extensible video stream multi code Rate of Chinese character multicast with all users is maximum turns to target function, take into account the interlayer dependence of extensible video stream decoding, with information flow equilibrium condition, network code condition, transmission node set time scheduling condition while of maximum, and chance routing mechanism is restricted to constraints for channel capacity, set up the dynamic code rate Resourse Distribute of telescopic video flowable state multi code Rate of Chinese character cast communication, Route Selection and transmitting and scheduling combined optimization problem.

Preferably, in 4th step, described complete distributed code rate allocation method refers to: use Duality Decomposition theory original optimization problem to be decomposed into some sub-optimization problems, each network node is allowed to utilize local local message adjust dynamically code check, transmitting and scheduling and Route Selection and upgrade, in a distributed way iterative thus realize the dynamic optimal Resourse Distribute of network.

More preferably, the concrete execution step of described complete distributed code rate allocation method is:

(a) initialization: arrange initial original/dual variable is some nonnegative value;

B () repeats:

For source node:

B1) dual variable information is received from destination node;

B2) dual variable information is received from next-hop node;

B3) from local storage, obtain the original of storage and dual variable information;

B4) original/dual variable is upgraded;

B5) to the dual variable information that set of network nodes transmission upgrades;

For each node:

B1) dual variable information is received from source node;

B2) dual variable information is received from next-hop node;

B3) from local storage, obtain the original of storage and dual variable information;

B4) original/dual variable is upgraded;

B5) to the dual variable information that the sending node transmission of this node upgrades;

B6) if this node is destination node, then to the dual variable information that source node transmission upgrades;

Until all original and dual variables converge to optimal value, or when reaching maximum iteration time, complete distributed code rate allocation method stops.

Compared with prior art, the present invention has following beneficial effect:

The needs of wireless network are become when the present invention is dynamic, provide a kind of complete distributed code rate allocation method, achieve dynamic networking Route Selection and transmitting and scheduling, and pass through the introducing of chance route and network code, improve the throughput of overall network and the adaptability to time varying channel states, for receiving terminal provides better video quality.

Accompanying drawing explanation

By reading the detailed description done non-limiting example with reference to the following drawings, other features, objects and advantages of the present invention will become more obvious:

Fig. 1 is the method flow diagram of one embodiment of the invention;

Fig. 2 is the schematic diagram of one embodiment of the invention multicast network, also show the basic module of chance route;

Fig. 3 is the schematic diagram of conflict relationship between transmission node in one embodiment of the invention network;

Fig. 4 is the schematic diagram of the distributed code rate allocation method of one embodiment of the invention;

Fig. 5 is the time scheduling schematic diagram of one embodiment of the invention to transmission node set while of maximum in network, and wherein (a) gives the relation curve that the link average data packet reception rate in network changed with each time period; Shown in (b), the optimal time scheduling result of transmission node set while of giving maximum in network.

Embodiment

Below in conjunction with specific embodiment, the present invention is described in detail.Following examples will contribute to those skilled in the art and understand the present invention further, but not limit the present invention in any form.It should be pointed out that to those skilled in the art, without departing from the inventive concept of the premise, some distortion and improvement can also be made.These all belong to protection scope of the present invention.

The present embodiment provides a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method based on chance route and network code, with reference to the method flow diagram shown in Fig. 1, comprises the steps:

1, the extensible video stream Data Rate Distribution constraint at source node place

As shown in Figure 2, the present embodiment carries out instance analysis to multicast network structure, can by abstract for wireless network be directed graph G=(V, E), wherein V is radio node set, and E is radio link set.Radio node set V can be expressed as three sub-union of sets collection further, i.e. V={s} ∪ N ∪ D, and wherein s represents source node, and N represents set of relay nodes, and D represents destination node set.Assuming that extensible video stream is encoded to M layer { L at source node place ₁, L ₂..., L _m, wherein the transmission code rate of m layer is positioned at tolerance interval in, and then use chance routing mechanism and network-encoding operation by multi-hop wireless via node by the extensible video stream multicast transmission of layering extremely each destination node.In addition, corresponding to the time-varying characteristics of wireless link, the whole time period of telescopic video flow transmission is evenly divided into several equal-sized time period t ∈ 1,2 ... T}, the channel status of wireless link remains unchanged within each time period.

2, the chance routing mechanism at via node place and network-encoding operation

As shown in Figure 2, also show the basic module of chance routing mechanism.Use chance routing mechanism, the wireless data packet sent by sending node transmission can be received by the multiple adjacent nodes in its effective transmission range.Owing to not specifying concrete down hop forward node, therefore any next-hop node successfully receiving this packet all likely continues to forward this packet.Therefore, chance routing mechanism be sending node select to receive packet and on geographical position than this sending node closer to the next-hop node of destination node as forward node set, and be dynamically that different packets selects actual down hop forward node, thus forwarded the broadcast characteristic that effectively make use of wireless shared media by local chance.

Assuming that in time period t, node s needs packet P ₁, P ₂and P ₃multicast transmission is to destination node d ₁~ d ₃.All down hop adjacent node n of node s ₁~ n ₆all in its effective transmission range, therefore can both receive these three packets sent by node s.But, in these six down hop adjacent nodes, owing to only having n ₁, n ₂and n ₃on geographical position than node s closer to destination node, the therefore set { n of this three nodes composition ₁, n ₂, n ₃be chosen as the forward node set of node s, be denoted as F _s(t).In fig. 2, connected node s and F is supposed further _st in (), the receives data packets probability of the link of each forward node is 2/3.Due between different receiving node wireless data packet receive time highly incoherent, therefore, assuming that three secondary data bags transmission after, the packet situation that each reception at Node arrives as shown in Figure 2, i.e. n ₁receive packet P ₁and P ₂, n ₂receive packet P ₁and P ₃, n ₃receive packet P ₂and P ₃.Then, by forward node set { n ₁, n ₂, n ₃the packet that receives will be forwarded to three destination nodes further by chance routing mechanism.Particularly, with destination node d ₂for example, consider F _s(t) interior joint and d ₂connectivity relation, node n ₂and n ₃the packet received is needed to be transmitted to d ₂.If node n ₂and n ₃between not do not link up, then these two nodes all expect forwarding data bag P ₃, thus likely cause destination node d ₂to packet P ₃repeat receive.In order to avoid this situation, traditional chance routing mechanism can based on node n ₂and n ₃to destination node d ₂space length distribute corresponding forwarding priority, if same packet is successfully received by multiple forward node, then select an actual forward node to continue to forward this packet according to the forwarding priority of these forward node.In this example, due to n ₂compare n ₃closer to destination node d ₂, n ₂forwarding priority be higher than n ₃.Therefore, if node n ₂successfully receive some packets, then this packet is forwarded to destination node d by being responsible for by it ₂, and node n ₃then avoid repeating transmission; Otherwise, if node n ₂do not receive this packet and node n ₃successfully receive this packet, then node n ₃by being responsible for, this packet is forwarded to destination node d ₂.Thus, node n ₂by forwarding data bag P ₁and P ₃to destination node d ₂, and node n ₃only forwarding data bag P ₂to destination node d ₂.Similarly, can by this forward node Resource selection and prioritization application of policies in the transmission of other two destination nodes.But, should be noted that the limitation of traditional chance routing mechanism needs to provide specific MAC protocol to determine the selection of forward node set and the cooperation transmission coordinated between this set interior joint.In addition, make the design of corresponding MAC protocol more complicated the multicast scene that traditional chance routing mechanism is applied to multiple destination node from clean culture scene.

By contrast, by introducing network-encoding operation, forward relay Nodes can perform algebraic operation operation to the packet received, thus overcomes two limitation of chance routing mechanism mentioned above.Such as, the packet directly forwarding and receive is different from, node n ₁the packet P after coding can be forwarded ₁+ P ₂to destination node d ₁, node n ₃the packet P after coding can be forwarded ₂+ P ₃to destination node d ₂and d ₃, and node n ₂still forwarding data bag P ₁and P ₃.Thus by using network-encoding operation in multicast scene, all three destination nodes can both successfully be recovered to obtain three raw data packets.That destination node no longer needs to receive all raw data packets by the benefit that chance routing mechanism combines with network-encoding operation, alternatively, the coded data packet (supposing that the number of raw data packets is K) receiving any K Line independent is only needed just successfully can to recover an original K packet.

3, the while of wireless network maximum, transmission node set divides and scheduling

As shown in Figure 3, be the conflict relationship figure in the original network topology structure shown in Fig. 2 between each transmission node.In the original network topology structure shown in Fig. 2 and telescopic video multicast network scene, node s, n ₁, n ₂and n ₃transmission node can be selected as send or forwarding data bag.But due to transmission node conflict, these four nodes can not carry out transmission or the forwarding of packet simultaneously.In the transmission node conflict relationship figure shown in Fig. 3, each end points represents a transmission node in original network topology, and has the one group of link set be connected with its forward node set.If two transmission nodes cannot carry out data packet transmission due to the conflict between its link be associated simultaneously, then there is conflict relationship between these two transmission nodes, corresponding to the line in conflict relationship figure between two end points.The concept of transmission node set is simultaneously introduced on this basis.When in transmission node set, all nodes carry out data packet transmission at the same time simultaneously, its link be associated can not produce wireless competition and can be successfully completed the data packet transmission of next-hop node.The basic thought of simultaneously transmission node set be by require in this set the route recipient that has an opportunity there is not at synchronization the conflict that competitive relation avoids between transmission node each other.In order to make full use of wireless shared media to realize the throughput upper limit of network, the while of defining maximum, transmission node set is a transmission node set simultaneously, if increase any one node in this set, causes this set to become nonsimultaneous transmission node set.

4, set up the dynamic network maximization of utility problem of many time periods based on chance route and network code, propose distributed code rate allocation method

Set up dynamic network maximization of utility problem of many time periods following (wherein the implication of each parameter can correspondingly within a context obtain):

Objective optimisation problems P1: $\underset{R,g,f,\mathrm{\λ}\≥0}{\max }\underset{t=1}{\overset{T}{\Σ}}\underset{d\∈D}{\Σ}\underset{m=1}{\overset{M}{\Σ}}U\left(R^{md}\right(t\left)\right)$

Constraints:

$1)---\underset{j:(i,j)\∈E}{\Σ}{g}_{(i,j)}^{md}\left(t\right)-\underset{j:(j,i)\∈E}{\Σ}{g}_{(j,i)}^{md}\left(t\right)={\mathrm{\σ}}_i^{md}\left(t\right);\∀i\∈V,m\∈M,d\∈D,t\∈T$

$2)---g_{(i,j)}^{md}\left(t\right)\≤f_{(i,j)}^m\left(t\right);\∀(i,j)\∈E,m\∈M,d\∈D,t\∈T$

$3)---\underset{\mathrm{\α}=1}{\overset{A}{\Σ}}{\mathrm{\λ}}^{\mathrm{\α}}\left(t\right)\≤1;\∀t\∈T$

$4)---\underset{j\∈J}{\Σ}\underset{m=1}{\overset{M}{\Σ}}{f}_{(i,j)}^m\left(t\right)\≤\underset{\mathrm{\α}=1}{\overset{A}{\Σ}}{\mathrm{\λ}}^{\mathrm{\α}}\left(t\right)\·C_{(i,J)}^{\mathrm{\α}}\left(t\right);\∀i\∈V,\∀J\⊆F_i\left(t\right),t\∈T$

$5)---\[\underset{t=1}{\overset{T}{\Σ}}{R}^{md}\left(t\right)\]\·\[\underset{t=1}{\overset{T}{\Σ}}{R}^{md}\left(t\right)-r_{\min }^m\·T\]\·\[\underset{t=1}{\overset{T}{\Σ}}{R}^{md}\left(t\right)-r_{\max }^m\·T\]\≤0;\∀m,d$

$6)---\frac{\underset{t=1}{\overset{T}{\Σ}}{R}^{(m+1)d}\left(t\right)}{r_{\max }^{(m+1)}}\≤\frac{\underset{t=1}{\overset{T}{\Σ}}{R}^{md}\left(t\right)}{r_{\min }^m};\∀m\∈\{1,2,...,M-1\},d\∈D$

Wherein, optimized variable is:

R-represents the vector be made up of video reception code check.Particularly, R ^mdt () represents that destination node d receives the code check of m video layer within t time period.

G-represents the vector be made up of virtual information stream bit rate.Particularly, represent virtual information stream bit rate wireless link (i, j) receiving m video layer within t time period for destination node d.

F-represents the vector be made up of actual bandwidth consumption amount.Particularly, represent within t time period on wireless link (i, j) for receiving the actual bandwidth consumption amount of m video layer.

λ-represent the vector be made up of time scale.Particularly, λ ^α(t) represent within t time period some specific maximum while transmission node set Г ^αt transmission time ratio that () is assigned to.

Optimization aim is:

The utility function summation of all users in wireless multicast network is maximized, namely maximizes the overall receiver, video quality of user in all T extensible video stream groupcast time section.Described user utility function U (R ^md(t)) represent that destination node d is with code check R ^mdt () successfully receives and the reduction of video distortion after m video layer of decoding.

Constraints is:

1) virtual information flow conservation condition, namely the inflow virtual information flow at arbitrary network node place is equal with outflow virtual information flow.Wherein, ${\mathrm{\σ}}_i^{md}\left(t\right)=\left\{\begin{array}{cc}{R}^{md}\left(t\right),& fori=s\\ -R^{md}\left(t\right),& fori=d\\ 0,& otherwise\end{array}\right..$

2) network code constraints, ensures that the bandwidth of actual consumption on wireless link is to the maximum in the virtual information flow of all destination nodes on this link.

3) the transmitting and scheduling condition of transmission node set while of maximum, be defined in any time can only have at most one maximum while transmission node set be arranged to transmit.

4) based on the radio transmission capacity constraints of chance routing mechanism and network-encoding operation, wherein, J is the forward node set F of node i _it a subclass of (), super limit (i, J) represents the link set of the broadcast area of the node composition in from node i to J, C _{(i, J)}be from node i to J in the effective transmission code rate of link of broadcast area of node composition.

5) and 6) be the interlayer dependency constraint condition of scalable video, guarantee each video coding layer of sequence reception that all destination nodes increase progressively from low to high successively according to label.

Distributed Problem Solving Algorithm and the implementation of above-mentioned optimization problem are as follows:

First the Lagrange duality form obtaining above-mentioned optimization problem is:

wherein, it is Lagrange multiplier.

Further, adopt that primal-dual interior pointmethod, upgrade original variable and dual variable, by iteration Step wise approximation optimum point, wherein k represents iterations, and δ is positive step value, [] simultaneously ⁺represent get on the occasion of computing:

$\begin{array}{c}{R}^{md}\left(t\right){|}_{k+1}=\[R^{md}\left(t\right)+{\mathrm{\δ}}^{\left(R\right)}\left(U^{\′}\right(R^{md}\left(t\right))+{\mathrm{\μ}}_s^{md}(t)-{\mathrm{\μ}}_d^{md}(t)-\frac{v^{(m-1)d}}{r_{\max }^m}+\frac{v^{md}}{r_{\min }^m}\\ -{\mathrm{\γ}}^{md}\{3{\[\underset{t=1}{\overset{T}{\Σ}}{R}^{md}\left(t\right)\]}^2-2T(r_{\min }^m+r_{\max }^m)\[\underset{t=1}{\overset{T}{\Σ}}{R}^{md}\left(t\right)\]+T^2{r}_{\min }^m{r}_{\max }^m\}){\]}_k^{+}\end{array}$

$g_{(i,j)}^{md}\left(t\right){|}_{k+1}={\[g_{(i,j)}^{md}\left(t\right)+{\mathrm{\δ}}^{\left(g\right)}\left({\mathrm{\μ}}_i^{md}\right(t)-{\mathrm{\μ}}_j^{md}(t)-{\mathrm{\η}}_{(i,j)}^{md}(t\left)\right)\]}_k^{+}$

$f_{(i,j)}^m\left(t\right){|}_{k+1}={\[f_{(i,j)}^m\left(t\right)+{\mathrm{\δ}}^{\left(f\right)}(\underset{d\∈D}{\Σ}{\mathrm{\η}}_{(i,j)}^{md}\left(t\right)-\underset{\underset{andj\∈J}{J\⊆F_i\left(t\right)}}{\Σ}{\mathrm{\β}}_{(i,j)}\left(t\right))\]}_k^{+}$

${\mathrm{\λ}}^{\mathrm{\α}}\left(t\right){|}_{k+1}={\[{\mathrm{\λ}}^{\mathrm{\α}}\left(t\right)+{\mathrm{\δ}}^{\left(\mathrm{\λ}\right)}\left(\underset{i\∈V}{\Σ}\underset{J\⊆F_i\left(t\right)}{\Σ}{\mathrm{\β}}_{(i,J)}\right(t\left)C_{(i,J)}^{\mathrm{\α}}\right(t)-\mathrm{\θ}(t\left)\right)\]}_k^{+}$

Wherein, the constraints that dual variable μ is corresponding is equality constraint, and therefore the renewal iteration of its correspondence does not need to carry out [] ⁺computing.

As shown in Figure 4, complete distributed code rate allocation method implementation following (wherein the implication of each parameter can correspondingly within a context obtain) is provided:

(a) initialization: arrange initial original/dual variable for some nonnegative value;

B () repeats:

For source node s:

B1 receives dual variable from destination node d

B2 receives dual variable from next-hop node j

B3 obtains the video reception code check R of storage from local storage ^md(t) | _k, dual variable v ^md| _k, γ ^md| _k, β _{(s, J)}(t) | _k, θ (t) | _ktransmission code rate effective in link

B4 upgrades and obtains video reception code check R ^md(t) | _k+1, virtual information stream bit rate actual bandwidth consumption amount transmission time ratio lambda ^α(t) | _k+1and dual variable

B5 to set of network nodes N ∪ D} transmit dual variable θ (t) | _k+1.

For each node i ∈ N ∪ D:

1. dual variable θ (t) is received from source node s | _k;

2. dual variable is received from next-hop node j

3. from local storage, obtain the dual variable of storage β _{(i, J)}(t) | _ktransmission code effective in link

4. upgrade and obtain virtual information stream bit rate actual bandwidth consumption amount transmission time ratio lambda ^α(t) | _k+1and dual variable

5. to { the sending node transmission dual variable in j ∈ V| (j, i) ∈ E}

If 6. node i is destination node, i.e. i=d, d ∈ D, then transmit dual variable to source node s

Until: all original and dual variables converge to optimal value, or till reaching maximum iteration time.

As shown in (a) in Fig. 5, give the relation curve that the link average data packet reception rate in network changed with each time period; As shown in (b) in Fig. 5, the optimal time scheduling result of transmission node set while of giving maximum in network.Wherein, corresponding with Fig. 2 and Fig. 3, three maximum while transmission node set be respectively Г ¹(t)={ s}, Г ²(t)={ n ₂and Г ³(t)={ n ₁, n ₃.Can see, because source node s is produced by scalable video by packet all the time, therefore the quality of no matter channel status, is Г within each time period ¹t optimal transmission ratio that () distributes remains larger value (about 0.4).On the other hand, when network condition is good time, receives data packets rate is relatively large, and the code rate allocation method proposed can for having transmission node (the such as Г of more multi-next-hop node ²node n in (t) ₂) distribute larger transmission time ratio; On the contrary, when network condition is bad time, receives data packets rate is relatively little, and the code rate allocation method proposed can improve utilance (the such as Г of other transmission node ³node n in (t) ₁and n ₃) so that many insecure wireless links are combined as a more reliable link.

The needs of wireless network are become when the present invention is dynamic, set up the dynamic network maximization of utility problem of many time periods based on chance route and network code, and accordingly provide a kind of complete distributed code rate allocation method, achieve dynamic extensible video stream Data Rate Distribution, network path selection and transmitting and scheduling, and pass through the introducing of chance route and network code, improve the throughput of overall network and the adaptability to time varying channel states, for receiving terminal provides better video quality.

Above specific embodiments of the invention are described.It is to be appreciated that the present invention is not limited to above-mentioned particular implementation, those skilled in the art can make various distortion or amendment within the scope of the claims, and this does not affect flesh and blood of the present invention.

Claims (10)

1. a telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method, is characterized in that, comprise the following steps:

The first step, at source node place, time-varying characteristics according to wireless network carry out segmentation to the time, within each time period, scalable video technology is used video flowing to be encoded to multiple scalable video layer, by ensureing the interlayer dependency constraint condition of scalable video to the Data Rate Distribution of different scalable video layer;

Second step, at via node place, when some concrete video datas wrap in and forward via the via node in network, use chance routing mechanism dynamically to select down hop forward node for it, and use network coding technique to improve the throughput of network further;

3rd step, in the wireless network, wireless network changes step response and node-node transmission competitive relation when having, each wireless network node in wireless shared media is divided into several maximum while transmission node set, realize the overall network transmission time each maximum while transmission node set between optimal scheduling;

The time variation step response and node-node transmission competitive relation of the chance routing mechanism that the 4th step, the interlayer dependent form condition combining the source node place video coding layer that three steps above obtain, via node forward and network coding technique and wireless network, uses the dynamic network maximization of utility modeling method of many time periods to realize the optimization problem of the telescopic video flowable state multi code Rate of Chinese character cast communication based on chance route and network code;

When carrying out concrete Data Rate Distribution scheme for terminal node, adopt complete distributed code rate allocation method, finally realize the combined optimization of the dynamic code rate Resourse Distribute of extensible video stream multi code Rate of Chinese character cast communication, dynamics route selection and dynamic transmission scheduling.

2. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1, it is characterized in that, in the first step, described time slice is the Further Division to extensible video stream groupcast time, in each time slice, the channel condition information of wireless channel is relatively stable, and the method passing through directly measurement and channel state prediction obtains.

3. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1 and 2, it is characterized in that, in the first step, the interlayer dependency constraint condition of described scalable video refers to: the decoding of the video layer that label is higher needs to depend on the low video layer of label, and namely the multicast transmission order of video layer needs to carry out according to video layer label order from low to high.

4. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1, it is characterized in that, in second step, described chance routing mechanism refers to: the broadcast characteristic and the space diversity characteristic that utilize wireless shared media, forwards in both candidate nodes set dynamically for packet selects actual down hop forward node in the down hop of via node.

5. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1 or 4, it is characterized in that, in second step, described network coding technique refers to: before via node forwards, carry out arithmetic coding operation to packet, and the actual bandwidth consumption amount specifying on every bar link is the maximum of all destination nodes bandwidth consumed on the link; This condition is the constraints adopting network code on link, to realize the resource-sharing on the same link of different destination node.

6. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1, it is characterized in that, in 3rd step, when while described, in transmission node set, all nodes carry out data packet transmission at the same time, its link be associated can not produce wireless competition and can be successfully completed the data packet transmission of next-hop node.

7. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1 or 6, it is characterized in that, in 3rd step, described maximum while transmission node set be a transmission node set simultaneously, if increase any one node in this set, cause this set to become nonsimultaneous transmission node set.

8. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1, it is characterized in that, in 4th step, described combined optimization refers to: the video total quality received in the complete period of extensible video stream multi code Rate of Chinese character multicast with all users is maximum turns to target function, take into account the interlayer dependence of extensible video stream decoding, with information flow equilibrium condition, network code condition, transmission node set time scheduling condition while of maximum, and chance routing mechanism is restricted to constraints for channel capacity, set up the dynamic code rate Resourse Distribute of telescopic video flowable state multi code Rate of Chinese character cast communication, Route Selection and transmitting and scheduling combined optimization problem.

9. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 1 or 8, it is characterized in that, in 4th step, described complete distributed code rate allocation method refers to: use Duality Decomposition theory original optimization problem to be decomposed into some sub-optimization problems, each network node is allowed to utilize local local message adjust dynamically code check, transmitting and scheduling and Route Selection and upgrade, in a distributed way iterative thus realize the dynamic optimal Resourse Distribute of network.

10. a kind of telescopic video flowable state multi code Rate of Chinese character multicast optimization transmission method according to claim 9, is characterized in that, described complete distributed code rate allocation method, and the concrete step that performs is:

(a) initialization: arrange initial original/dual variable is some nonnegative value;

B () repeats:

For source node:

B1) dual variable information is received from destination node;

B2) dual variable information is received from next-hop node;

B3) from local storage, obtain the original of storage and dual variable information;

B4) original/dual variable is upgraded;

B5) to the dual variable information that set of network nodes transmission upgrades;

For each node:

B1) dual variable information is received from source node;

B2) dual variable information is received from next-hop node;

B3) from local storage, obtain the original of storage and dual variable information;

B4) original/dual variable is upgraded;

B5) to the dual variable information that the sending node transmission of this node upgrades;

B6) if this node is destination node, then to the dual variable information that source node transmission upgrades;

Until all original and dual variables converge to optimal value, or when reaching maximum iteration time, complete distributed code rate allocation method stops.