CN104243098A - Transmitting method based on parallel batch network coding - Google Patents
Transmitting method based on parallel batch network coding Download PDFInfo
- Publication number
- CN104243098A CN104243098A CN201410492165.5A CN201410492165A CN104243098A CN 104243098 A CN104243098 A CN 104243098A CN 201410492165 A CN201410492165 A CN 201410492165A CN 104243098 A CN104243098 A CN 104243098A
- Authority
- CN
- China
- Prior art keywords
- batch
- source node
- transmission
- freedom
- scheduling
- 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.)
- Pending
Links
Abstract
The invention relates to a transmitting method based on parallel batch network coding. According to the transmitting method, network coding is used for transmitting data, data flow to be transmitted is divided into batches comprising the same number of groups, the groups in each coding batch are coded and transmitted together, and a target node can decode one batch after receiving the enough number of the coded groups. The method can be applied to a wireless Mesh network, and provides reliable and efficient data transmission service; network coding is utilized for improving the transmission efficiency; to overcome the defect of low efficiency caused by an existing research and use stop wait strategy, the invention provides a parallel transmission method; a sending window which can dispatch a plurality of batches at the same time is maintained on a source node, and parallel dispatching of the multiple batches is controlled according to feedback information; in comparison with an existing research result, batch transmission can be organized more flexibly and efficiently, and then the end-to-end handling capacity is improved.
Description
Technical field
The present invention relates to a kind of method transmitting data in wireless Mesh netword, on network code (Network Coding) basis, this method uses a parallel batch of scheduling, transmission mechanism, can organize a batch transmission more flexibly, efficiently, and then improve end-to-end throughput.The invention belongs to field of wireless communications networks.
Background technology
Wireless Mesh netword is the broadband multi-hop wireless network that a class is positioned Large Copacity, two-forty, and it has the features such as accurate static, energy constraint is weak, in network capacity, reliability, robustness etc., demonstrate huge advantage.In wireless broadcast channel, multiple nodes that transfer of data all can be transmitted in scope each time receive, and therefore, the wireless network comprising Wireless Mesh network presents significant information redundancy.Under traditional " Store and forword " transmission means, if it is not that oneself grouping directly will abandon it that node receives destination node, the information redundancy that radio broadcasting characteristic causes cannot be effectively utilized.If reasonably can utilize the broadcast characteristic of wireless channel, the performance of wireless Mesh netword will have improves space greatly.
Network code gives good thinking to how utilizing the broadcast characteristic of wireless channel, when the multiple nodes in transmission range receive one and same coding grouping, different node therefrom can obtain different, required separately information, thus effectively can improve the utilization ratio of radio channel resource.In addition, the information that individual data divides into groups is carried in multiple coding groups by application network coding, can also reduce the dependence of transfer of data to single grouping, thus improves the efficiency of transmission in unreliable wireless environment further.
When using network code, data to be transmitted stream is divided into batch (Batch) different batches comprising the grouping of identical number and is coded separately transmission, and destination node is decodable code one batch 0 after receiving the coding groups of enough numbers.Here a key issue is: how source node organizes the transmission of multiple batches, while guarantee transmission reliability, obtain high efficiency of transmission.For this reason, existing research uses and stops waiting strategy, and before receiving the acknowledge message of destination node, source node continues the coding groups 0 sending present lot] [3] [4].Existing method can ensure transmission reliability but efficiency is not high, source node after the transmission completing batch, receive the confirmation message before, will a large amount of useless redundancy packets be sent, thus cause the decline of efficiency of transmission.Especially, in remote (jumping figure is large) communication scenes, end-to-end feedback delay can be very large, thus can cause and now methodically sharply to worsen.
To sum up, needing a kind of transmission method being applicable to wireless Mesh netword badly, making full use of network code in the wireless network while advantage, improving efficiency of transmission further.
[list of references]
[1]P.A.Chou,T.Wu,K.Jain,Practical?network?coding,in?Proc.of?Allerton?Conference,pp,40-49,2003.
[2]S.Chachulski,M.Jennings,S.Katti,et?al,Trading?structure?for?randomness?in?wireless?opportunistic?routing,in?Proc.of?ACM?SIGCOMM,pp.169-180,2007.
[3]J.Subramanian,R.Morris,H.Balakrishnan,UFlood:High?throughput?flooding?over?wireless?mesh?networks,in?Proc.of?IEEE?INFOCOM,pp.82-90,2012.
[4]X.Zhang,B.Li,Optimized?multipath?network?coding?in?lossy?wireless?networks,IEEE?JSAC,vol.27,no.5,pp.622-634,2009.
[5]J.K.Sundararajan,D.Shah,M.Médard,ARQ?for?network?coding,in?Proc.of?IEEE?ISIT,pp.1651-1655,2008.
Summary of the invention
Technical problem: the present invention relates to a kind of transmission method based on parallel batch network code, utilize network code, this method, in conjunction with a parallel batch of transmission, scheduling mechanism, effectively can reduce the dependence of transfer of data to feedback, and then improve end-to-end efficiency of transmission.
Technical scheme: the present invention uses network code to transmit data, data to be transmitted stream is divided into comprise identical number grouping batch, grouping in each coding batch is coded in be transmitted together, and destination node to receive after the coding groups of enough numbers i.e. decodable code one batch.
Thought of the present invention can be illustrated as: the transmission that source node uses a send window to safeguard multiple batches simultaneously, dispatches concurrently according to feedback information and batch state, sends multiple batches.More specifically:
● source node is dispatched and after sending one batch, is waited for that destination node feeds back the accepting state of this batch, and other batch during this period in scheduling window;
● source node knows the state of each batch by feedback information, if destination node is successfully decoded one batch, source node should batch to be removed from send window, and put into one new batch; If destination node fails decoding one batch, source node will dispatch this batch again;
● to be scheduled batch for each, the encoded packet transmission amount of source node to it depends on the accepting state of this batch and the source node quality of connection to destination node.
One batch will be scheduled once or repeatedly, until source node receives the confirmation decoded information of destination node.This method can alleviate the dependence of transfer of data to feedback information, under the prerequisite not losing transmission reliability, improve efficiency of transmission.
Transmission method based on parallel batch network code of the present invention is specially:
1) network code transmitting data information is used, source node data to be transmitted stream is divided into comprise identical number N grouping batch, grouping in each coding batch is coded in be transmitted together, be decodable code one batch after destination node received code grouping number >N, N is linear independence coding groups number;
2) transmission of multiple batches safeguarded by the send window that source node uses can hold the memory space of some batches simultaneously, after source node scheduling transmission one batch, wait for that destination node feeds back the reception condition of this batch, the i.e. degree of freedom, and dispatch all the other batches in send window during this period;
3) source node knows the accepting state of each batch by feedback information, if destination node is successfully decoded one batch, source node should batch to be removed from send window, and put into one new batch; If destination node fails decoding one batch, source node will dispatch this batch again;
4) source node according to batch state determine batch dispatching sequence and opportunity, batch dispatching priority is followed successively by from high to low: the new lot be not scheduled, be fed back to after scheduling do not decode batch, do not obtain after scheduling feedback batch; For equal state batch, the less priority of batch sequence number is higher, and feedback time more early priority is higher;
5) the source node encoded packet transmission amount that determines to be scheduled batch according to quality of connection between batch reception condition and source, destination node, to be scheduled batch for each, encoded packet transmission amount is that the degree of freedom of this batch is divided by the degree of freedom delivery rate of source node to destination node.
The account form of described degree of freedom delivery rate is: whenever source node receives a batch of b
kaccepting state feedback, source node calculate nearest one take turns scheduling since b
kdegree of freedom reduction Δ d
k, and inquire about nearest one and take turns in scheduling and send b
kthe number n of coding groups
k, degree of freedom delivery rate is updated to Δ d
k/ n
k.
Beneficial effect: the present invention relates to a kind of method transmitting data in wireless Mesh netword, while making full use of network code, can alleviate the negative effect of feedback to efficiency of transmission.By safeguarding that at source node can be dispatched the send window of multiple batches simultaneously, and the Parallel Scheduling of multiple batches is controlled according to feedback information, a batch transmission can be organized more flexibly, efficiently: be just removed send window after 1) being only identified decoding when one batch, it is hereby ensured transmission reliability; 2) source node is while wait one batch feedback, can dispatch all the other batches, therefore, it is possible to improve the level of resources utilization, and improves throughput.
Accompanying drawing explanation
Below in conjunction with accompanying drawing, the present invention is further described.
Fig. 1: functional module composition and more intermodule relation schematic diagram,
Fig. 2: batch state transfer schematic diagram.
Embodiment
Before detailed process describes, first introduce two definition:
Definition 1 (degree of freedom, Degree offreedom): the number of the coding groups of the linear independence that node decoder one batch also needs
[5].
Definition 2 (degree of freedom delivery rates, DDR): the ratio of the coding groups number that the linear independence coding groups number (i.e. degree of freedom reduction) that destination node receives and source node send, when using network code for reflecting, source node is to the quality of connection of destination node.
In this method, functional module composition and each intermodule correlation are shown in Fig. 1.
Specific implementation of the present invention is:
1) network code transmitting data information is used, source node data to be transmitted stream is divided into comprise identical number N grouping batch, grouping in each coding batch is coded in be transmitted together, be decodable code one batch after destination node received code grouping number >N, N is linear independence coding groups number;
2) transmission of multiple batches safeguarded by the send window that source node uses can hold the memory space of some batches simultaneously, after source node scheduling transmission one batch, wait for that destination node feeds back the reception condition of this batch, the i.e. degree of freedom, and dispatch all the other batches in send window during this period;
3) source node knows the accepting state of each batch by feedback information, if destination node is successfully decoded one batch, source node should batch to be removed from send window, and put into one new batch; If destination node fails decoding one batch, source node will dispatch this batch again;
4) source node according to batch state determine batch dispatching sequence and opportunity, batch dispatching priority is followed successively by from high to low: the new lot be not scheduled, be fed back to after scheduling do not decode batch, do not obtain after scheduling feedback batch; For equal state batch, the less priority of batch sequence number is higher, and feedback time more early priority is higher;
5) the source node encoded packet transmission amount that determines to be scheduled batch according to quality of connection between batch reception condition and source, destination node, to be scheduled batch for each, encoded packet transmission amount is that the degree of freedom of this batch is divided by the degree of freedom delivery rate of source node to destination node.
The account form of described degree of freedom delivery rate is: whenever source node receives a batch of b
kaccepting state feedback, source node calculate nearest one take turns scheduling since b
kdegree of freedom reduction Δ d
k, and inquire about nearest one and take turns in scheduling and send b
kthe number n of coding groups
k, degree of freedom delivery rate is updated to Δ d
k/ n
k.
1 source node operation:
1.1 send window
1. data to be transmitted stream is divided into isometric batch by source node, and each batch comprises M original packet;
2. source node uses a send window (size is W batch) to transmit multiple batches simultaneously, and determines order and the opportunity of dispatching lot according to feedback information;
3., after source node dispatches one batch, wait for that destination node feeds back the accepting state of this batch, and dispatch all the other batches in send window at the same time;
4. source node knows the accepting state of each batch by feedback information, if destination node is successfully decoded one batch, source node should batch to be removed from send window, and put into one new batch; If destination node fails decoding one batch, source node will dispatch this batch again.
1.2 batches of scheduling
Note batch b
kstate be S
k, batch state comprises to be waited for dispatch state (WF_SCH) and waits for feedback states (WF_FBK), and state transitions is shown in Fig. 2.Note b
ktime of being scheduled the last time be T
k, T
kcan be used for differentiation one batch whether to be scheduled, if T
k=-1 (initial value), then b
kbatch (being also called new lot) that be not scheduled.The present invention by following principle scheduling send window batch:
1. first, consider to dispatch be not scheduled new lot (exist multiple satisfy condition batch time scheduling sequence number reckling), sending coding groups number is the degree of freedom/DDR (estimating see 1.3 quality of connection: DDR estimates) to be scheduled batch, upgrades the scheduling time of corresponding batch;
2. secondly, consider scheduling be in WF_SCH state all the other batches (exist multiple satisfy condition batch time scheduling T
kreckling), namely accepting state known batch, sending coding groups number is the degree of freedom/DDR to be scheduled batch, upgrades the scheduling time of corresponding batch;
3. dispatch WF_FBK state batch, because source node not yet grasps the up-to-date accepting state of this kind of batch, for the single batch of too much coding groups of transmission has the risk causing too much redundant transmission arbitrarily, therefore, randomized policy is adopted to be dispersed in multiple batches by coding transmission, batch b
kexpectation encoded packet transmission amount and its wait scheduling time T
k, degree of freedom d
kbe directly proportional;
1.3 quality of connection are estimated: DDR estimates
1. (b might as well be supposed whenever source node receives one batch
k) accepting state feedback, source node calculate nearest one take turns scheduling since b
kdegree of freedom reduction Δ d
k, and inquire about nearest one and take turns in scheduling and send b
kthe number n of coding groups
k, may be calculated Δ d according to definition 1, DDR
k/ n
k;
2., in order to alleviate random perturbation, make the measured value upgrading DDR with the following methods:
Wherein DDR ' is current measurement value,
for forgetting factor, Δ d
kless forget slower.
1.4 encoding operation
Source node does stochastic linear coding to current transmission batch, and structure coding groups is as follows
Wherein, P
1~ P
mfor the grouping of current scheduling batch, α
1~ α
mfor the code coefficient from Stochastic choice in a finite field gf (q), whole addition and multiply operation are all in the upper execution of GF (q).α=[α
1..., α
m] be called as coding vector, be embedded into coding groups stem and data division together transmits.
2 destination node operations
2.1 decode operation
1. when node receives a coding groups, use Gaussian elimination (Gaussian Elimination) judge this grouping whether with buffered packet linear independence, buffer memory linear independence grouping (also claiming the grouping of " newly " (Innovative)), abandons linear correlation grouping;
2. represent the linear independence grouping number of batch of destination node buffer memory with m, as m=M, destination node is by Gaussian elimination and matrix inversion decoding respective batch.
2.2 send feedback
1. destination node obtain decoder can time send an acknowledgment packet, and in this feedback the degree of freedom of incidentally all the other batches;
2. destination node safeguards a feedback timer, if exceed given length apart from the last time sending feedback packet, then sends a feedback packet.
Claims (2)
1., based on a transmission method for parallel batch network code, it is characterized in that:
1) network code transmitting data information is used, source node data to be transmitted stream is divided into comprise identical number N grouping batch, grouping in each coding batch is coded in be transmitted together, be decodable code one batch after destination node received code grouping number >N, N is linear independence coding groups number;
2) transmission of multiple batches safeguarded by the send window that source node uses can hold the memory space of some batches simultaneously, after source node scheduling transmission one batch, wait for that destination node feeds back the reception condition of this batch, the i.e. degree of freedom, and dispatch all the other batches in send window during this period;
3) source node knows the accepting state of each batch by feedback information, if destination node is successfully decoded one batch, source node should batch to be removed from send window, and put into one new batch; If destination node fails decoding one batch, source node will dispatch this batch again;
4) source node according to batch state determine batch dispatching sequence and opportunity, batch dispatching priority is followed successively by from high to low: the new lot be not scheduled, be fed back to after scheduling do not decode batch, do not obtain after scheduling feedback batch; For equal state batch, the less priority of batch sequence number is higher, and feedback time more early priority is higher;
5) the source node encoded packet transmission amount that determines to be scheduled batch according to quality of connection between batch reception condition and source, destination node, to be scheduled batch for each, encoded packet transmission amount is that the degree of freedom of this batch is divided by the degree of freedom delivery rate of source node to destination node.
2. the transmission method based on parallel batch network code according to claim 1, is characterized in that the account form of described degree of freedom delivery rate is: whenever source node receives a batch of b
kaccepting state feedback, source node calculate nearest one take turns scheduling since b
kdegree of freedom reduction Δ d
k, and inquire about nearest one and take turns in scheduling and send b
kthe number n of coding groups
k, degree of freedom delivery rate is updated to Δ d
k/ n
k.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410492165.5A CN104243098A (en) | 2014-09-23 | 2014-09-23 | Transmitting method based on parallel batch network coding |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410492165.5A CN104243098A (en) | 2014-09-23 | 2014-09-23 | Transmitting method based on parallel batch network coding |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104243098A true CN104243098A (en) | 2014-12-24 |
Family
ID=52230524
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410492165.5A Pending CN104243098A (en) | 2014-09-23 | 2014-09-23 | Transmitting method based on parallel batch network coding |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104243098A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104580009A (en) * | 2015-01-28 | 2015-04-29 | 华为技术有限公司 | Chip data forwarding method and device |
| CN106330399A (en) * | 2016-08-30 | 2017-01-11 | 中国人民解放军理工大学 | Rapid feedback method applicable to network coding of multihop heterogeneous network |
| CN107040334A (en) * | 2015-09-30 | 2017-08-11 | 香港中文大学 | Loss for communication network restores agreement |
| CN109104262A (en) * | 2018-07-03 | 2018-12-28 | 深圳市智慧海洋科技有限公司 | A kind of wireless reliable communication method of the underwater sound |
| CN109547552A (en) * | 2018-11-29 | 2019-03-29 | 浙江鲸腾网络科技有限公司 | API request processing method, device, storage medium and electronic equipment |
| CN110248384A (en) * | 2019-06-20 | 2019-09-17 | 中国科学院大学 | Single-hop wireless multicast block length optimization method based on random linear network encoding |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101667885A (en) * | 2009-09-29 | 2010-03-10 | 天津大学 | Method for reducing redundancy message by using network coding technique on DTN or ICN network |
| WO2010085212A1 (en) * | 2009-01-21 | 2010-07-29 | Agency For Science, Technology And Research | Method and system for linear network coding in broadcast systems |
| CN102547587A (en) * | 2010-12-28 | 2012-07-04 | 上海贝尔股份有限公司 | Method and device for carrying out data multicasting in batches in wireless multicasting system |
| CN102970117A (en) * | 2012-12-05 | 2013-03-13 | 中国人民解放军理工大学 | Method for suitable for end-to-end degree of freedom feedback of random linear network coding |
| CN103250463A (en) * | 2010-11-23 | 2013-08-14 | 香港中文大学 | Subset coding for communication systems |
-
2014
- 2014-09-23 CN CN201410492165.5A patent/CN104243098A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010085212A1 (en) * | 2009-01-21 | 2010-07-29 | Agency For Science, Technology And Research | Method and system for linear network coding in broadcast systems |
| CN101667885A (en) * | 2009-09-29 | 2010-03-10 | 天津大学 | Method for reducing redundancy message by using network coding technique on DTN or ICN network |
| CN103250463A (en) * | 2010-11-23 | 2013-08-14 | 香港中文大学 | Subset coding for communication systems |
| CN102547587A (en) * | 2010-12-28 | 2012-07-04 | 上海贝尔股份有限公司 | Method and device for carrying out data multicasting in batches in wireless multicasting system |
| CN102970117A (en) * | 2012-12-05 | 2013-03-13 | 中国人民解放军理工大学 | Method for suitable for end-to-end degree of freedom feedback of random linear network coding |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104580009A (en) * | 2015-01-28 | 2015-04-29 | 华为技术有限公司 | Chip data forwarding method and device |
| CN104580009B (en) * | 2015-01-28 | 2018-08-21 | 华为技术有限公司 | Chip forwards the method and device of data |
| CN107040334A (en) * | 2015-09-30 | 2017-08-11 | 香港中文大学 | Loss for communication network restores agreement |
| CN106330399A (en) * | 2016-08-30 | 2017-01-11 | 中国人民解放军理工大学 | Rapid feedback method applicable to network coding of multihop heterogeneous network |
| CN106330399B (en) * | 2016-08-30 | 2019-05-07 | 中国人民解放军理工大学 | A kind of fast feedback approach suitable for multi-hop heterogeneous network network code |
| CN109104262A (en) * | 2018-07-03 | 2018-12-28 | 深圳市智慧海洋科技有限公司 | A kind of wireless reliable communication method of the underwater sound |
| CN109547552A (en) * | 2018-11-29 | 2019-03-29 | 浙江鲸腾网络科技有限公司 | API request processing method, device, storage medium and electronic equipment |
| CN109547552B (en) * | 2018-11-29 | 2022-01-25 | 恒生电子股份有限公司 | API request processing method and device, storage medium and electronic equipment |
| CN110248384A (en) * | 2019-06-20 | 2019-09-17 | 中国科学院大学 | Single-hop wireless multicast block length optimization method based on random linear network encoding |
| CN110248384B (en) * | 2019-06-20 | 2023-05-09 | 中国科学院大学 | Single-hop wireless multicast packet length optimization method based on random linear network coding |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104243098A (en) | Transmitting method based on parallel batch network coding | |
| US8432848B2 (en) | Queued cooperative wireless networks configuration using rateless codes | |
| US7710908B2 (en) | Method and arrangement for coding and scheduling in a retransmission communication system | |
| KR20020042438A (en) | Hybrid ARQ with parallel packet transmission | |
| CN102255712B (en) | Soft storage allocation method for hybrid automatic repeat request process of long term evolution system | |
| CN103051424B (en) | A kind of radio transmitting method of unequal error protection fountain codes | |
| Cam et al. | Throughput analysis of some ARQ protocols in the presence of feedback errors | |
| CN101286819B (en) | Data receiving method and device | |
| CN101911569A (en) | Be used for keeping the technology of service quality of the connection of wireless communication system | |
| US20150333874A1 (en) | A transmitting and a receiving device and respective methods performed thereby for improving retransmissions in wireless networks | |
| CN101540778B (en) | Method, device and system for transmitting data | |
| Akin et al. | Backlog and delay reasoning in HARQ system | |
| JP2018510531A (en) | Method and system for on-demand file repair | |
| CN101562506B (en) | Data transmission method | |
| CN101459490A (en) | Data transmission method and device | |
| CN105391518B (en) | A kind of document transmission method towards deep space network | |
| CN104869543A (en) | Information transmission method and device for eNodeB and user equipment | |
| Sagduyu et al. | On broadcast stability region in random access through network coding | |
| Malak et al. | ARQ with cumulative feedback to compensate for burst errors | |
| Zhang et al. | Probabilistic network coding with priority over wireless networks | |
| Sorour et al. | A network coded ARQ protocol for broadcast streaming over hybrid satellite systems | |
| CN110958078B (en) | Low-delay stream code packet transmission method for high-loss link | |
| CN100452713C (en) | Network data concast transmission method based on distributed coding storage | |
| Enenche et al. | Network Coding as Enabler for Achieving URLLC Under TCP and UDP Environments: A Survey | |
| Shao et al. | Hash searching and network coding based constant retransmission for wireless multicast |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20141224 |