The application requires in the priority of the U.S. Provisional Application the 61/096th, No. 570 (title is " Generic Retransmission Scheme forCommunication Systems ") of submission on September 12nd, 2008.
The application also requires in the priority of the U.S. Provisional Application the 61/096th, No. 636 (title is " Flexible Layer Retransmission Scheme UsingCorrelation Information at different Layers ") of submission on September 12nd, 2008.
The application also requires the U.S. Provisional Application the 60/976th of U.S.'s non-provisional application 2007P52589 number, submission on October 2nd, 2007, the U.S. Provisional Application the 60/984th that on October 31st, No. 839 1 submitted to, the priority that No. the 60/991st, 812, the U.S. Provisional Application of submitting in No. 132 and on December 3rd, 2007.
The application also requires the U.S. Provisional Application the 60/976th of U.S.'s non-provisional application 2007P52629 number, submission on October 2nd, 2007, the U.S. Provisional Application the 60/984th that on October 31st, No. 808 1 submitted to, the priority that No. the 60/991st, 809, the U.S. Provisional Application of submitting in No. 162 and on December 3rd, 2007.
Content interim and non-provisional application listed above is hereby expressly incorporated by reference.
Embodiment
Now, describe one or more execution mode of the present invention with reference to the accompanying drawings, wherein, similarly reference number is used for the expression similar elements.Although the example of re-transmission scheme hereinafter is discussed under the environment of VDSL and ADSL system, is should be noted that the present invention generally speaking can be applicable to any communication system.
Fig. 1 shows the embodiment of DSL communication system 100.As is known to the person skilled in the art, DSL communication system 100 can be DMT (Discrete multi-tone) system, and wherein, data are modulated on a plurality of subcarriers, makes each subcarrier be associated with a carrier frequency.DSL communication system 100 comprises the first transceiver 102a (such as switch), rack or other network temination units that is arranged in the operating point 104.The first transceiver 102a is connected to the second transceiver 102b via subscribers feeder 106.The second transceiver 102b is integrated in customer premises equipment, CPE (CPE) subscriber unit 108 (such as also being integrated in such as the modem/router in other equipment of personal computer or notebook or other gateways arbitrarily).
The first transceiver 102a comprises the first transmitter 112a and the first receiver 114a that is connected to subscribers feeder 106.The second transceiver 102b comprises the second transmitter 112b and the second receiver 114b that is connected to subscribers feeder 106.In order to connect transmitter and receiver, each transceiver all can comprise the connecting interface such as hybrid network etc.
The first controller 110a can be set to control and to coordinate the function of the first transceiver 102a.In addition, can the second controller 110b be set to control and to coordinate the function of the second transceiver 102b in user's side.Though Fig. 1 shows respectively first and second controller 110as, the 110b integrated with the first and second transceiver 102a, 102b, should be appreciated that the first and second controller 110a, 110b can be arranged with each transceiver branch.In addition, should be understood that parts such as first and second controllers all can comprise a plurality of parts and can realize with hardware, software, firmware or its combination in any.
In addition,, should be appreciated that the first transceiver 102a can be connected to a plurality of subscriber units 108 though Fig. 1 only shows a subscribers feeder to the long-distance user, each can have a plurality of second transceiver 102b.In addition, in certain embodiments, can be in conjunction with the subscribers feeder more than two so that higher data rate to be provided to the user.
In order to understand how swap data of DSL communication system 100 better, explain some layers of the example network protocol stack 200 of VDSL or ADSL system with reference to Fig. 2.Fig. 2 shows minimum two-layer in the osi model, that is, and and data link layer 202 and PHY layer 204.For clear and simplification, the high level in the osi model is not shown.According to Fig. 2, PHY layer 204 is divided into three PHY sublayers.
The one PHY sublayer is PMD (physical medium dependence) layer 206, and it comprises such as symbol and regularly generating and recovery, Code And Decode, modulation and demodulation, echo are eliminated the basic function of (if being implemented) and line conditioning, link startup and physical layer 204 (superframe) etc.In addition, pmd layer 206 can generate or receiving control message via the expense passage.
Next PHY sublayer is PMS-TC (physical medium appointment-transmission convergence) layer 208, and it is connected to pmd layer 206 by δ interface (delta interface).PMS-TC layer 208 is chain of commands, and the management entity in CO and CPE modulator-demodulator provides management basic indication.PMS-TC layer 208 also provide such as the generation of frame and frame synchronously, the function of scrambling (descrambling), Reed-Solomon coding and interpolation.
The 3rd PHY sublayer is TPS-TC (host-host protocol appointment-transmission convergence) layer 210.It is connected to PMS-TC layer 208 by the α interface (Alpha's interface) that is in central office website place or the β interface (beta interface) that is in user site.TPS-TC layer 210 provides such as the tissue of packing framing, Bearer Channel, multiplexed etc. function.TPS-TC layer 210 is connected to data link layer 202 by γ interface (gamma interface).
Along with data are handled by network protocol stack 200 and are transmitted by subscribers feeder 106, impulsive noise and crosstalk noise can be used the data of transmitting unfriendly.The data that this noise meeting deterioration is transmitted make loss of data or hiding efficient communication.In the existing xDSL system as ADSL and VDSL, have designated some mechanism (as Trellis coding, RS coding and interpolation) to alleviate this noise effects.Yet along with popularizing (wherein, the information of losing can cause " flicker " at video screen) such as the service of audio frequency popular, expectation provides the service quality than the better grade of service quality that can be by the current techniques realization.
Repeating transmission is a kind of technology that proposes in this application, in order to increase video and other application quality by DSL.In certain embodiments, retransmitting function is inserted in the data link layer 202 or physical layer 204 of network protocol stack 200.In the solution formerly, retransmitting function is " gamma interface " or " Alpha's interface " peculiar (therefore, retransmitting function cannot exchange between layer).On the contrary, in some embodiments of the invention, retransmitting function is independent of the particular network stack layer.
Fig. 3 shows general retransmission protocol 250.As shown in the figure, the first transceiver 102a is sent to data flow for the network protocol stack of transmitting at 252 places.At 254 places, the first transmitter 112b transfers to the second transceiver 102b by subscribers feeder 106 with data streams.In case the second transceiver 102b receives institute's data flow transmitted, the second controller 110b with regard to the data cell in the specified data stream whether at 256 places by deterioration.If detect the data cell of deterioration, the second transceiver 102b just requires at 258 places to retransmit by the data cell ie of deterioration.At 260 places, the first transceiver 102a comes this re-send request may is responded by retransmitting the data cell of asking, thereby makes the second transceiver 102b recover original data stream at 262 places.Should be appreciated that in other embodiments, the second transceiver 102b can be used as transmitter/repeating transmission machine, and the first transceiver 102a can be used as receiver.Below with reference to remaining accompanying drawing more detailed embodiment is described.
Referring now to Fig. 4 A~Fig. 4 B, can see more detailed retransmission protocol.Particularly, Fig. 4 A shows from the transmitter (agreement of) viewpoint for example, the first transceiver 102a, and Fig. 4 B shows from the agreement of the viewpoint of receiver (for example, the second transceiver 102b).This embodiment is based on being grouped into some data cells in the container, then with container identifier (CID) and container to related as the reference that is used for retransmitting.Compare with prior art system, this embodiment has realized low overhead rate, causes more efficiently communication.
In emitting side, Fig. 4 A shows the payload data stream 300 that receives at network interface layer 303 (for example, at the gamma interface).Payload data stream 300 comprises a plurality of each payload data unit 302 (PL) that are suitable for retransmitting (EL) or are unsuitable for retransmitting (NEL) that are denoted as.In certain embodiments, each data cell 302 constitutes single complete packet or unit (for example, Ethernet data bag or asynchronous transfer mode (ATM) unit) according to the network protocol layer of better grade.
Each data cell can be made up of payload header 304 and payload data 306.Payload header 304 generally includes payload sequence identifier (PLSID) 308 and retransmission identifier 310 (NEL/EL).PLSID 308 designated data unit in payload data stream 300 with respect to the position of other data cells, thereby make the upper-layer protocol at receiver place with the suitable order payload data stream that divides into groups again.Retransmission identifier 310 designation data unit are suitable for retransmitting (EL) and still are unsuitable for retransmitting (NEL).For example, the real-time voice data can be classified as NEL, because repeating transmission can cause the unacceptable stand-by period (delay) between the session participant.On the contrary, audio frequency or FTP data can be classified as EL, can not cause the unacceptable stand-by period because it can be buffered.
After network interface layer 303 places received, this payload data stream 300 was handled by container controller 312 (can be positioned at the gamma interface).The container controller is grouped into payload data stream 300 in the container flow of being made up of a series of containers 314 (for example, C1, C2, C3, C4).Each container comprises container header and a series of data cell.For example, container 4 (C4) comprises container header (CH4) and four payload data unit: payload EL-13, payload NEL-14 and two additional/pseudo-units (pad/dummyunit).Although Fig. 4 A (for example shows data cell that all containers all comprise equal number, four data unit) example, but in other embodiments, different containers can comprise the data cell of varying number, and is just passable as long as transmitter and receiver is known different length.
Container header 316 can comprise a plurality of fields.For example, shown field comprises: (a) data in the expression container header are launched in the very first time or the retransmitted first emission/retransmission field (RTX); (b) identify the container identifier (CID) of given container uniquely; (c) embed overhead channel (EOC) byte; (d) (#EL) field of the EL data cell quantity in the expression container; And the reserved field that (e) can comprise other useful informations.
After generating container, DTU controller (data transmitter unit controller 318 can be positioned at Alpha/beta interface) container handling, thus generate DTU stream (data transmitter unit stream 320).((for example, C1), redundant digit (for example, R1) and embed overhead channel (EOC) byte (for example, EOC1) for example, DTU1) all to comprise single container for shown each DTU.Under the situation of ATM, redundant digit is based on Reed-Solomon (RS) check code; And under the situation of ATM, redundant digit is based on Cyclical Redundancy Check (CRC).In other embodiments, each DTU comprises a plurality of containers, and different DTU can comprise the container of varying number.
The data streams 324 that mission controller 322 forms by subscribers feeder 106 transmission.This data streams can comprise the container of transmission for the first time and retransmit container.When transmission data streams 324, the data among noise (for example, impulsive noise) the meeting deterioration DTU, for example, represented by " X " on DTU1 and the DTU3.
If receiver detects misdata, then receiver is sent back to the first transceiver 102a with re-send request may 326.Re-send request may 326 is specified the container that utilizes misdata to receive.For example, in the embodiment shown, the RRC field 328 in the re-send request may 326 can specify in the CID of the deterioration container of receiver place reception.In certain embodiments, RRC field 328 can be attached the upstream payload data (USPL) 330 that transfers to the first transceiver 102a from the second transceiver 102b, and EOC field 332 and redundant digit 334.In one embodiment, re-send request may 326 comprises the byte of every character fixed number.In this case, receiver can be asked the container number of every character.
In case receive re-send request may 326, retransmit controller 336 and can recover the container of execution in RRC field 328.In order to be beneficial to this function, in retransmitting buffer 338, stored the single copy (the perhaps container that is associated with DTU at least) from each DTU of data streams 324.These DTU can be stored in and retransmit in the buffer 338 the certain time expiration after transmission or guidance repeating transmission buffer 338 is filled with.In recovering the process of retransmitted container, retransmit controller 336 and can use the CID of expression in re-send request may 326 look-up table relevant with capsule address in the repeating transmission buffer 338.
After searching retransmitted container, retransmit controller 336 and only from original container, find out payload EL data (namely, ignore payload NEL data), add new data (EL, the NEL or the additional data that for example, before do not have transmission) then to fill up the remainder with retransmitted container.Therefore, in the example of Fig. 4 A, reinstate original EL data cell 1, original EL data cell 2 and original EL data cell 3 with the data cell one of new data and retransmit container 1 (C1)
*And reinstate original EL data cell 9 with two data unit one of new data and original EL data cell 10 is retransmitted container 3.Be used for these container header of retransmitting container and comprise the CID of original container, and specify this container for retransmitting container.Then, add redundant digit and EOC byte to container, and will retransmit container and be inserted in the data streams and by subscribers feeder 106 and be transmitted.
Now, with reference to Fig. 4 B, can see the receiver side agreement.In this example, receiver receives DTU stream by subscribers feeder 106, and it comprises by the DTU of deterioration (DTU1, DTU3).Then, data receiving element (DRU) controller 350 uses redundancy bytes whether to detect in each DTU storage errors.If possible, DRU controller 350 can use redundancy bytes to proofread and correct these mistakes.In case receive container flow (comprising " hole " that the deterioration container takes place now), container controller 352 is just noticed to be needed to retransmit container, and sends control signals to re-send request may controller 354.For example, in Fig. 3, control signal can specified containers 1 and container 3 by deterioration, therefore, should be retransmitted.Then, re-send request may controller 354 is sent back to the first transmitter 102a with re-send request may 326, and it is retransmitted to indicate which container not need.At last, receiver receives the container C 1 of the repeating transmission that can also comprise new data
*, C3
*, and will recover the payload data stream 300 of original transmitted.
A subject matter of some existing re-transmission schemes is not exist to embedding the protection of overhead channel (EOC) data.In order to address this problem, in certain embodiments, in case receive re-send request may, transmitter just checks whether the CID of container is associated with the EOC data.If there are the EOC data that are associated with container, then the EOC data are retransmitted to receiver with the payload EL data of container.
Fig. 5 shows container controller 312 and how data cell is grouped into another embodiment in the container.In this example, container is indicated CID (can reside in the container header) then by the container size of the adjacent byte of EL data packing up to maximum.If before reaching the largest container size, run into the NEL data, then indicate CID after " closure " container and the direct EL data behind fringe.Then, packing NEL data in next container.This packaging technique makes the receiver request only comprise the repeating transmission of the container (that is, institute's request container does not comprise the NEL data) of EL data.Then, DTU is grouped to comprise continuous container as shown in the figure, and is added with redundant and EOC information.Therefore, for very long adjacent data bag (for example, video packets of data, this is typical case), this embodiment has realized being used for the low-down expense retransmitted.
Now, with reference to Fig. 6, can see another embodiment of the re-transmission scheme that is suitable for dissimilar TC layer payload.Yet present embodiment has some similar to previously discussed, and in this embodiment, the payload data unit is the Ethernet data bag that is divided before in being grouped into container.
More specifically, receive the stream of Ethernet data bag 500 at segment controller 502 places.Each Ethernet data bag comprises payload data 504 and ethernet header (EH) 506, and it specifies the Ethernet data packet number.Then, segment controller 502 is divided into segment 508 with the Ethernet data bag, and to each fragment additional clips sequence identifier FSID.
Then, container controller 510 wraps in segment 508 in the container 512, thereby can " divide " some segments between container.For example, fragment 3 is divided between container C 1 and C2.Usually, each container can comprise the slice header (FH) of segment boundaries in container header (CH) and a series of distinguish containers.As following more detailed discussion, FH can comprise FSID and other information.
After container was divided into groups, DTU controller 514 added the Reed-Solomom redundancy bytes end of each container to, to form data streams 516.Then, by circuit transmission data streams.
If receiver receives the data that can't proofread and correct, then receiver can will the FSID corresponding with the data that can't proofread and correct offer the repeating transmission controller of transmitter by circuit.Then, retransmit controller and can retransmit the necessary data that is associated with FSID.Note, in this embodiment, because the DSL system is kept for carrying out the various counters of monitoring, so as long as transmitter and receiver keeps the counting of DTU, then do not need to transmit any label except FSID.Therefore, because transmitter and receiver are synchronous in DSL, so in this embodiment, do not need CID is transferred to receiver from transmitter.
As shown in Figure 7, section headers (FH) comprises FSID field and segment end (EF) field.Shown in the EF field comprise shown encoding scheme, it can be used for identification segment end and how to aim at respect to container ends.Also can use other encoding schemes.Added field (not shown) in the slice header can also be indicated in the given segment or the EL data between the segment and the aligning of NEL data, perhaps can indicate in the given segment or the retransmission data between the segment and the new aligning of launching data.
In one embodiment, the container with EL data has unique Container ID, and has the ID that the container of NEL has previous appropriate containers.With the repeating transmission adequate information and the unique serial ID that are used for data cell, repeater system can use the error detection/correction technology to identify the deterioration data, and whether identification is belonged to suitable data flow by the data of deterioration.An advantage of this scheme is to have only proper container just to carry unique ID, and the ID of unsuitable data carrying appropriate containers.So, mend suitable data capsule if lose, then receiver will can not be asked the data of losing, because must there be the previous container that had successfully received with proper data.
Fig. 8 shows the method 700 of repeating transmission, briefly discusses now.Though show mode 700 and be described as a series of actions or event, should be appreciated that the present invention is not limited by this action shown in the lock or the order of event.For example, according to the present invention, can some actions take place with different order, and/or be accompanied with the action of and/or description shown with this paper or other actions or the event that event is separated.In addition, the not all action that illustrates or event all need according to the invention process method.
In 702, transmitter comprises the data streams of a series of containers by the transmission medium emission.
In 704, receiver receives the data streams that can change by the noise on the transmission medium.Receiver is identified the data flow that receives by the wrong identification information of transmitting in the evaluation data streams and whether is comprised the deterioration container.
In 706, suppose to find the deterioration container, then, transmit re-send request may from receiver to transmitter.Re-send request may is specified during the transmission by one or more containers of noise degradation.
In 708, re-send request may is launched machine and handles.In this step, carry out look-up table so that the container of asking with relevant by the data cell of original transmitted in institute's request container.In 710, in next used vessel, retransmit the data cell in institute's request container.
In certain embodiments, can use " container copy " to replace " frame blanking ", thereby in the time interval that take place to repeat electric pulse noise (REIN), stop temporarily or limit transmission.
Therefore, above-described embodiment is the flexible re-transmission scheme of restricting data cell size not.This re-transmission scheme uses the expense of " just enough " as required, because have only proper data compressed.Can understand, as long as there is the mixing of suitable and inappropriate data, above-mentionedly suitably can cause additional overhead, but from using and there is not this situation in practice situation, because the long data bag normally adjacent as the proper data of video.
Although described above-described embodiment (that is, two suitable grades are can represent by unit), but in other embodiments, can comprise the additional level of suitable grade about suitable/unaccommodated data cell for retransmitting.For example, can use suitable grade more than three.For example, in certain embodiments, " grade 1 " can be to hang down (for example to retransmit grade of fit, only retransmit container No. one time), " grade 2 " can be that middle grade (is for example retransmitted grade of fit, only retransmit container twice), and " grade 3 " can be high-grade repeating transmission grade of fit (for example, just retransmitting twice container correctly to send desired message when needing).Can use multiple standards to be fit to retransmit or be not suitable for repeating transmission so that data cell is categorized as, example includes but not limited to: applied, suitable down in specific external condition (for example, noise); Always be not suitable for; Always be fit to; And/or suitably in time window, retransmit.Can be based on the proper standard that dynamically changes the data class that is associated with repeating transmission as the external perimysium reference of noise characteristic.For example, if system detects impulsive noise on the line, then can be set at " grade 1 " with retransmitting grade of fit, and if system detects white Gaussian noise on the line, then can be set at " grade 3 " with retransmitting grade of fit.Although the example with 3 grades has been discussed, should be appreciated that, according to performance requirement, this grade of fit class of emission can be expanded to infinitely in practice.
Although illustrate and described the present invention with reference to one or more execution modes, under the situation of the spirit and scope that do not deviate from claims, can change and/or revise example illustrated.For example, although described the present invention with reference to the ADSL and the VDSL communication system that communicate by a pair of pair of twisted copper wire, the present invention can be applicable to the transmission medium of any communication system and any type.For example, other communication systems can comprise cell phone, pager, mobile communication equipment, industrial control system, wide Local Area Network, local area network (LAN) etc.These and other systems can communicate by various types of communication mediums, include but not limited to unlimited medium, optical fiber, coaxial cable, power line etc.
In addition, although each illustrated embodiment is depicted as hardware configuration, function of the present invention and corresponding characteristic can also be carried out by the combination of appropriate software Lu You or hardware and software.Implement for software, term as used herein " computer-readable medium " refers to participation provides arbitrary medium from instruction to the controller (for example, microprocessor) that installs or be associated with device.This medium can adopt arbitrary form, includes but not limited to non-volatile media, Volatile media and transmission medium.For example, non-volatile media comprises CD or disk.Volatile media comprises dynamic memory, for example SRAM or DRAM.Transmission medium comprises the inner or outside bus of coaxial cable, copper cash, optical fiber and device.Transmission medium can also comprise electromagnetic wave, such as electricity, light wave or radio wave.
Particularly, about the various functions of being carried out by above-mentioned parts or structure (assembly, device, circuit, system etc.), unless otherwise specified, term (comprising the reference to " device ") that be used for to describe these assemblies corresponding to any parts of the formulation function of the execution parts of describing or structure (for example is used for, function equivalent), though on the structure with the disclosed structure inequivalence of carrying out the function of the exemplary enforcement of the present invention shown in this paper.In addition, though only disclose concrete property of the present invention with respect to a kind of in a plurality of execution modes, this feature can combine with one or more other characteristics of other execution modes, can expect for the advantage of given or application-specific arbitrarily.In addition, term " comprises ", " having " or their distortion are used in that details is described and claim in, this term is similar to the mode that term " comprises ".