CN1939020A - Communication system - Google Patents

Communication system Download PDF

Info

Publication number
CN1939020A
CN1939020A CN200580010803.4A CN200580010803A CN1939020A CN 1939020 A CN1939020 A CN 1939020A CN 200580010803 A CN200580010803 A CN 200580010803A CN 1939020 A CN1939020 A CN 1939020A
Authority
CN
China
Prior art keywords
data
substation
central station
described central
parameter group
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN200580010803.4A
Other languages
Chinese (zh)
Other versions
CN1939020B (en
Inventor
保罗·安东尼·德尔夫
拉塞尔·保罗·戴维
大卫·布赖恩·佩恩
安德鲁·洛德
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
British Telecommunications PLC
Original Assignee
British Telecommunications PLC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from GB0502257A external-priority patent/GB0502257D0/en
Application filed by British Telecommunications PLC filed Critical British Telecommunications PLC
Priority claimed from PCT/GB2005/001252 external-priority patent/WO2005096574A1/en
Publication of CN1939020A publication Critical patent/CN1939020A/en
Application granted granted Critical
Publication of CN1939020B publication Critical patent/CN1939020B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Optical Communication System (AREA)

Abstract

The present invention relates to a communication system having a central station and a plurality of outstations. The central station is configured to execute a compensation procedure for compensating for degradation of data from the outstations, the compensation procedure having a plurality of adjustable characteristics, which characteristics are governed by a parameter set. The central station stores a respective parameter set in respect of each outstation; and, for each outstation, applies the compensation algorithm to data from that outstation using the parameter set associated with that outstation. In an initialization phase, the central station tests a plurality of starting parameter sets on data from a given outstation and selects the set providing the best compensation. The selected set is stored for subsequent compensation of data from that outstation.

Description

Communication system
Technical field
The present invention relates to central station, particularly, relate to a kind of central station that the deterioration from the data of a plurality of substations is compensated of being configured to.
Background technology
The compensation program (routine) that is used for distortion correcting signal is known.Yet when the source of the data that arrive central station changed on the short time yardstick, perhaps when arriving the serious distortion of data, this known program was not always suitable.
Summary of the invention
According to an aspect of the present invention, a kind of central station that is used for receiving from a plurality of substations data is provided, described central station is configured to carry out in use and is used for compensation process that the deterioration from the data of described a plurality of substations is compensated, described compensation process has at least one tunable characteristic by parameter group control, wherein, described compensation process may further comprise the steps: (i) use different start-up parameter groups that the data from substation are compensated; (ii) the quality of using the data after described different start-up parameter group compensates is measured; And, select a start-up parameter group that the data from the follow-up arrival of described substation are compensated (iii) according to the quality of being measured.
In this way, can test, thereby can select suitable initial parameter group to come the data on flows that arrives is compensated a plurality of start-up parameter groups.This will make that a station can reach apace can be to using the state of gratifying parameter group from the data of given substation.
Preferably, in order to carry out meticulousr adjusting to described compensation process, described compensation process has a plurality of tunable characteristics, and described parameter group comprises a plurality of parameters that are used to control described a plurality of tunable characteristics respectively.
Preferably, described central station is configured to store a parameter group at each substation.This can reduce described central station all calculated or selected to be used for the new parameter value of this substation when the data that at every turn receive from substation needs, make described central station easier, especially arrive under the situation of described central station with burst (shortburst) form in data from any given substation to compensating from any distortion in the arrival data of substation.Described parameter this locality can be stored in described central station place, perhaps can described parameter be stored in the remote location place by described central station.
Preferably, according to above step (i) to (iii) selecting at each substation stored parameters group.Preferably, when receiving the data of the substation that connects of making a fresh start at described central station place, carry out described selection at each substation at initial phase.Can store at from the data of given substation and selected group of selecting, so that at carrying out follow-up use from the data of this substation, this follow-up use may occur in and receive and/or compensated after the data from different substations.Yet, for example can use the adaptive equalization process that the group of stored parameter that is used for given substation is improved, the value after storage improves is so that use at carry out the later stage from the data of this substation.
Can come from other start-up parameter groups, to select a start-up parameter group by to using the quality and the predetermined quality threshold value that do not realize on the same group to compare.Yet, preferably,, the quality through offset data (that is, the tolerance of quality) of using different startup groups to realize selects described selected group by being compared, and described selected group is the group that has realized first water.
Preferably, the data that are used for described different parameters group (at least at given substation) is compared are identical data, can duplicate this data at described central station place, for use in different start-up parameter groups.
Preferably, use and from the test data of substation the start-up parameter group is assessed, preferably, the copy with this test data before the test data from substation arrives is stored in described central station place.This makes and can compare the test data of being stored and the test data through compensation from substation, thereby can assess the quality through offset data.Described different start-up parameter group test data can be applied to concurrently, perhaps alternatively, each parameter group these data can be applied to successively.
Described compensation process can also may further comprise the steps: from a plurality of time locations place in the data flow of substation described data flow is being sampled; And to each sample execution respective function, preferably, each respective function is controlled by relevant parameters or parameter group.Described respective function can correspond respectively to a characteristic of described compensation process.Each function can be simple weighting function, and each parameter is a weight coefficient.Yet these parameters can be controlled other functions that can be applicable to all sample points.For example, comprise at described compensation process under the situation of Fourier transform step, these parameters can be respectively applied for Fourier coefficient is weighted.
Preferably, described a plurality of substations are provided for sending data, so that arrive described central station continuously: promptly from the data of different substations, make under the situation that has adopted wavelength division multiplexing, non-intersect folded from the data of different substations, the data that send on same frequency channels are non-intersect folded at least.Described substation can one by one send data pulse string (being preferably numerical data) by circulation or other sequential systems.Preferably, press broadcast mode by described central station and send scheduling and other timer commands, when send data to indicate described a plurality of substation.
Provide substation when can send prenoticing of data owing to be stored in the dispatching command at described central station place, therefore can be chosen in Incoming (incoming) to described central station place employed parameter group when data are used described compensation process (, described dispatching command can be determined which parameter group of selection or which parameter at least in part) easily according to institute's stored scheduling instruction.
These substation transmitted traffic data of permission that described dispatch command can comprise at the appointment substation reach the set period instruction of (the continuous period or the period of segmentation).Thus, preferably, dispatch command will allow substation to send limited amount data.For example, upstream flowrate can be arranged to the stream of a plurality of frames, each frame all comprises a plurality of cells or a plurality of other sub-portions, and schedule information can indicate or allow, and branch stands in frame that one or more is specified in cell and sends data.In case substation responds dispatch command, preferably, this substation is the dispatch command (although not needing dispatch command just can retransmit timing data or other management datas) of products for further before sending further data on flows just.Therefore, in case sent data from substation as the result of first dispatch command, preferably, this substation is the dispatch command of products for further before sending further data on flows just.
In one embodiment, described central station has a plurality of continuous stream partial data streams with reception, each stream part all has the different labels that are associated with it, at least some stream parts have and the different distorted characteristic of other stream parts, wherein said central station can be operated with access scheduling information, can infer the respective labels that joins with Incoming stream part correlation according to this schedule information.Described central station subsequently can: the label of inferring the data flow part according to this schedule information, and select to be used for the stream parameter group partly that is associated with the label of being inferred according to the label of being inferred.
Preferably, described compensation process comprises adaptive algorithm, and this adaptive algorithm is used for when the data that receive from given substation the value of described parameter being improved.Described a plurality of substations can be configured to be sent in the known data in described central station place, so that described adaptive algorithm is trained.When only partly described adaptive algorithm being trained, only can store and preserve by in the improved described parameter of part some or all, so that when same substation receives further data, using after a while from the data of given substation.Therefore, preferably, with described central station be constructed so that when the source that arrives data when first substation changes over second substation, described central station: (i) storage is at the improvement value from the data of first substation, so that retrieve after a while; (ii) retrieval is before at the second substation stored parameters.
Train (except being used to select the start-up parameter group) in order to help described adaptive algorithm, described central station can storing predetermined data, and described or each substation can be stored corresponding tentation data, described or each substation are configured to its tentation data is sent to described central station, make and to compare tentation data of being stored and the tentation data that is received at described central station place.Preferably, how be corrected in order to make described adaptive algorithm assess distortion data more easily, at least some in the described data can be identical (between substation and described central station, different substations can have different training datas).Then, described adaptive algorithm can compare the data after proofreading and correct and the former state data before the distortion when initialization (that is, data).Can during fabrication this " known " data be included in the permanent memory chip of described central station and a plurality of substations.
Preferably, the optical-fiber network that is provided with such as fiber optic network is connected to each substation with described central station.Yet, also can between described central station and described a plurality of substation, adopt radio communication.If employing optical-fiber network, then this network preferably has at least one branch's contact (being also referred to as " splitter (splitter) "), be not multiplexed on the common carrier so that can have the seedbed, thereby these signals flow to as time multiplexing date and reach described central station from the signal of a plurality of substations.Thus, preferably, this communication network is configured to, at the flow that propagates into described central station from described substation (promptly, " upstream " flow) carry out work in time division multiple access (TDMA) mode, each minute stands in the transmission phase and sends thus, and the timing that sends the phase is in response to from one or more command instruction of described central station and definite, transmission to each substation is selected, to reduce the risk that overlaps mutually or arrive simultaneously described central station from the data of different substations.
Description of drawings
In appended independent claims, describe other aspects of the present invention in detail.Only the present invention is described in more detail referring now to accompanying drawing in the mode of example, in the accompanying drawing:
Fig. 1 shows the communication system that has central station and a plurality of substations according to of the present invention;
Fig. 2 a and 2b respectively illustration be positioned at the broadcasting and the staggered multiplexing transmission at branch contact place in the communication system of Fig. 1;
Fig. 3 a and 3b show downstream and upstream frames form respectively;
Fig. 4 schematically shows the functional representation of the central station in the communication system of Fig. 1;
Fig. 5 schematically shows the functional representation of one of a plurality of substations shown in Figure 1;
Fig. 6 (i) is to (iv) showing the data pulse string that arrives central station from a plurality of substations;
Fig. 7 is the table that the step in the backoff algorithm is shown;
Fig. 8 a and 8b show the different lines of the table of a plurality of steps in the expression adaptive compensation algorithm together;
Fig. 9 a and 9b show expression together and are used for handling different lines from the table of a plurality of steps of another adaptive algorithm of the data of different substations;
Figure 10 shows the flow chart of the key step with Fig. 9;
Figure 11 shows the alternative of backoff algorithm and represents; And
Figure 12 a and 12b show from other examples of the data of a plurality of substations arrival;
The flow chart illustration of Figure 13 the algorithm that wherein different cranking system arrays is compared with parallel mode;
The flow chart illustration of Figure 14 the algorithm that wherein different cranking system arrays is compared with serial mode; And
Figure 15 (i), (ii) show when data being tested different cranking system array compensation to the data pulse train.
Embodiment
Fig. 1 shows optical-fiber network 10, and it is also referred to as EPON (PON), and wherein, central station 12 (being also referred to as optical line terminal or OLT) is connected to a plurality of substations 14 (each substation is also referred to as optical network unit or ONU) by fiber optic network 16.Fiber optic network 16 comprises trunk optical fiber part 18, is connected to this trunk optical fiber part 18 in a plurality of branch optical fiber parts 20 in contact 21 places that formed by coupler or splitter.In the branch optical fiber part 20 each can have coupled phase Sub Ammunition Supply Poi 14.In addition, some or all in these fiber sections can have coupled corresponding other couplers 21, to be connected to a plurality of other components.
Contact 21 is provided so that the light intensity that propagates into substation 14 (this direction is called " downstream " direction) from central station 12 sentences preferred evenly mode at contact 21 and be distributed in the middle of branch or the sub-branch's optical fiber.In opposite or " upstream " direction, that is, on the direction of central station, do not have the seedbed at contact 21 places and close from the light of branch optical fiber.Usually, EPON comprises a plurality of contacts and sub-contact, and each contact n has 32 " along separate routes " usually,, joins 32 branches of public trunk to that is.
In order to communicate with substation, central station sends broadcast on downstream direction, and this broadcast usually can be received by all substations, these message have in the expression substation which or which be the recipient's of expection label or identifier.
Fig. 2 a illustration the transmission on the downstream direction.In updrift side, continuous a plurality of branches stand in separately the time slot 201 and send data, these time slots be provided so that when from the light of different branch optical fibers at contact 21 places and fashionable, should not overlap mutually from the data of different substations.In this way, it is staggered or to be equal to the ground time division multiplexing be frame structure 202 not have the seedbed at the contact place from the data of substation, as illustrated in Fig. 2 b.Central station 12 visits the data from this station by read time slot (that is, using time division multiple access (TDMA) agreement) from each station.In PON, use ATM (asynchronous transfer mode) cell between substation and central station, to communicate usually.The example of the frame structure with a plurality of ATM cell has been shown among Fig. 3 a (at downstream direction) and Fig. 3 b (at updrift side).The length of ATM cell is 424, and protected frequency band separates on updrift side, to consider the timing error from the transmission that each substation carries out.Except that the data payload, each ATM cell all contains the head of the information that comprises addressing field and other define styles.On downstream direction, these frames have isolated signaling cell, and the ATM cell carrying is present in the data on flows between these signaling cells.In the example of Fig. 3 b, the signaling cell is the PLOAM cell that is used for distributing (or be equal to ground, authorize) bandwidth by the OLT cell, that is, and and " physical layer operation, management and maintenance " cell.The PLOAM cell also provides synchronously, and is used for range finding (ranging), ERROR CONTROL, fail safe and other " maintenance " functions.The speed that sends the PLOAM cell can be limited and can be changed according to the requirement of OLT and substation (ONU) by central station (OLT).Can use upstream field in the ATM cell to realize signaling along updrift side.
Illustrate in greater detail central station 12 among Fig. 4.Central station 12 comprises: network input stage 40 is used for receiving data by common fiber (trunk optical fiber) from substation; Network output stage 41 is used for sending data by common fiber or along another optical fiber that downstream direction transmits to substation; Be connected to the central controller 42 of input stage 40 and output stage 41, be used for the moment of each substation transmission data is controlled; Compensating module 44 is used for the data that receive from substation are handled, and compensates or is equal to ground " equilibrium " with any distortion that may suffer data; Data output stage 46 is used for and will exports to recipient's (not shown) from described substation through offset data; And the data input stage, be used for data to be sent are inputed to described substation.
Realize central controller 42 and compensating module 44 at least one processor facility 50 with at least one processor that is used for deal with data.In addition, central controller 42 and compensating module 44 all can be visited and be used to store memory of data facility 52.Memory facility 52 can be distributed between central controller and the compensating module, and compensating module can be visited local RAM memory 521 and/or rapid access memory 522.This memory facility 52 can comprise pre-determined factor memory 523.Central controller can also comprise and being used for regularly and synchronous clock 43, and the buffer 45 that is used for the data that are sent to network are carried out short-term storage.
Network input stage 40 has photodetector 55, the light signal that this photodetector 55 is used for receiving converts the signal of telecommunication to, make compensating module in electric territory, to handle received light signal, preferably, compensating module is implemented on one or more electrical chip device.Be provided with the laser 57 (or other light sources) that is used to generate the light signal that is sent to substation at output stage 41 places, light signal is modulated, perhaps generate light signal in response to the signal of telecommunication that comes self-controller 42.
Illustrate in greater detail substation 14 among Fig. 5.This substation has: input stage 60 is used for from central station 12 receiving optical signals; Light compensated stage 62 is used for the possible distortion from the signal of central station along fiber path is compensated; Interface level is used for (i) and receives flow from least one user terminal, to send it on the optical-fiber network and (ii) the flow that this optical-fiber network was transmitted is sent to user terminal; Output stage 66 is used for flow is sent to optical-fiber network; And timing level 68, be used for controlling via the timing of the flow of interface level 64 from the local user.Input stage 60 comprises photodetector 61, and this photodetector 61 is used for converting light signal to the corresponding signal of telecommunication, thus the compensated stage of substation 62 can be in electric territory processing signals.
Regularly level 68 is connected to interface level 64, sends to flow on the network 16 to receive expection.Give output stage with the flow of correct timing as electrical signal transfer, this output stage has the laser aid 67 (or other sources) that is used for flow is become with other electrical signal conversion light signal.Thereby regularly level 68 can be carried out timing to the transmission of signal according to the instruction that receives from central station, preferably, should the timing level be connected to input stage by compensated stage 62.Particularly, regularly level 68 comprises clock 69 and the buffer 70 that is used for before sending data being carried out short-term storage.In order to make substation carry out work, can be provided with the processor 75 with at least one processor and storage arrangement 77, this memory and processor are implemented on one or more chip apparatus.
Under a kind of mode of operation, the central controller 42 of central station is configured for broadcasting frame structure with the signaling cell part of PLOAM cell (for example as), this signaling cell comprises: frame synchronizing signal; Be used to indicate selected substation to return the dispatch command that cell sends accordingly by one or more, perhaps be equal to ground, be used to indicate and select substation sends data according to the phase delay with respect to the reception of frame synchronizing signal instruction.Central controller 42 with dispatch command be stored in memory 52 (preferably, but the part of the fast access of this memory, as its RAM part) in.This can make the miscellaneous part (as compensating module) of this system can utilize these dispatch commands to infer the sign of substation (receiving data from this substation).
Preferably, the clock 69 at each substation place is configured to, with the data synchronised in the broadcast frame such as cell synchronizing signal or other signals (may be scrambled data).So the timing level 68 of substation 14 just can be used from the signal of this clock delay-level is controlled, make before sending data, to make the receive delay fixed time of data with respect to frame synchronizing signal.As a result, central controller 42 can send the data from substation and carry out " scheduling ".Can influence (affect) this scheduling according to designated period (this cycle can be variable), especially adopt under the situation of Dynamic Bandwidth Allocation, be like this through regular meeting in course of normal operation.
As distributing the alternative means of the situation of a window, controller can indicate a substation to send data at initial phase to each substation and indicate time that all the other branches stand in specified amount not to avoid sending data with interior, up to sent instruct till.Thus, central station can exchange a series of message with given substation, and other substations do not participate in.
In order to take into account from the different transit times (transit time) of different substations, central station will be realized " range finding " process, in this ranging process, the selected substation backhaul signals of central station indication reaches the time of specified amount, or backhaul signals is (alternatively immediately after receiving instruction, central station can indicate selected divide to stand in receive the time that backhaul signals after the cell synchronizing signal reaches specified amount, this cell synchronizing signal sends from substation by constant duration usually).According to institute's elapsed time between instruction transmission and the backhaul signals arrival, central controller can calculate the differential delay skew, and this deviant is sent to this substation.These deviants of the timing level 68 of this substation storage then make this deviant is included in any future from central controller in the time-of-the-day order.In this way, the data that sent by substation will be according to sending with the corresponding additional delay of this deviant.
In order to keep the timing level 68 of each substation can being configured to send the sync check signal synchronously by the time interval.Preferably, after receiving frame synchronizing signal, send the inspection signal from each substation, or opposite feasible inspection signal from each substation arrives central station by corresponding assigned address in returning frame by corresponding specified delay.Central controller 42 stores described assigned address and the sign of corresponding substation or the mapping that address correlation gets up of each being checked signal.Thus, if do not receive the inspection signal in position expeced time of returning frame, then controller can be given an order to corresponding substation, this order comprises the expression to the time difference between the time of reception of checking signal and the expection constantly, the timing level of substation responds this order, so that delay-level 70 will postpone to change and represented corresponding amount of time difference.In this way, can make asynchronous (out-off-time) substation get back to synchronous regime.
Compensating module 44 is configured to move backoff algorithm, and this backoff algorithm is used for data are handled or equalization, to proofread and correct this data at any inter symbol interference (ISI) or other distortions.This backoff algorithm has a plurality of tunable characteristics, and each tunable characteristic is all represented an aspect of the mode that this backoff algorithm handles data.Control these tunable characteristics by one group of coefficient (parameter), make each coefficient all be associated with corresponding characteristic.Selection to the value of these coefficients will be depended on the degree of distortion and the character of distortion (type).The character of distortion may depend on a plurality of factors, and for example distance, the material behavior of optical fiber, the bit rate propagated along optical fiber of signal also may depend on the residing environmental aspect of optical fiber.
Typically, for existing P ON, substation and central station fiber distance at interval is about 20km, but wishes to be at least 50km, 100km, 150km or even the fiber lengths of 200km.Distance between central station and the substation usually according to substation and different, difference is 20km, 50km or even 100km.Except being convenient to the distortion of (owing to different substation distance causes) sudden change with because the distortion of the increase that causes of the absolute distance of growing compensates, the present invention's (at corresponding substation place) the also easy to use light source with wider bandwidth (more cheap such as the source of the wider bandwidth of Fabry-Perot (FabryPerot) laser than source such as the narrow bandwidth of Distributed Feedback Laser, but can cause the higher distortion factor).
Particularly, should be appreciated that in the context of the present invention that the nature and extent of distortion may be different for the data from different substations.In addition, although the distortion of going into office data at central station place may change (that is, fast as about per 424 bits) according to cell, for the data from given substation, if there is distortion, this distortion may be to change lentamente.
Compensating module has been stored corresponding coefficient sets at each substation with the form of table, this table has comprised a group identifier in delegation, identifier of each substation, and in another row, comprised corresponding coefficient sets, this table has comprised the map information of each substation maps identifiers to one group of coefficient.According to institute's stored scheduling information, compensating module is inferred the sign of the substation that cell of data was derived from of each arrival.Infer that at every turn data will be when different substations arrive, compensating module all retrieves the coefficient sets at this substation storage, and moves described algorithm according to the coefficient that newly retrieves, so that the data from current substation are handled.Thus, as the result that scheduling is handled, compensating module has known in advance which substation next cell to be received will be derived from.Therefore compensating module can switch between suitable coefficient fast, compensates with the distortion that next is arrived cell.As a result, adjacent cell can have very different amount distortion, because backoff algorithm needn't be at the optimization process that starts anew of each cell, but can adjust back suitable coefficient sets apace, has therefore kept suitable value in backoff algorithm.
Preferably, this backoff algorithm is adaptive, that is, preferably, this backoff algorithm has since one group of initial value the value of one group component is carried out improved optimizer.In the case, when the source that arrives data when first stop changes over second station, compensating module is configured to: (i) storage is at the improvement value that calculates from the data of first substation, so that the later stage is retrieved; (ii) retrieve or the previous coefficient that had calculated at second substation of readjustment.Preferably, the dispatch command of origin self-controller provides execution in step (i) and triggering signal (ii), and these dispatch commands will indicate in the future arrival data with the sign of the substation that is derived from advance.Thus, whenever the source of received data will be when a substation changes over another substation, central controller can provide signal.Alternatively, compensating module can (in local random access memory 521) only storage representation receive (generating) schedule information in the expection source of data by central controller, or visit this schedule information on the contrary.
Can be with reference to Fig. 6 (i) to the operation of (iv) illustrating compensating module, Fig. 6 (i) each in (iv) all shows from the distortion of the signal of substation A, B, C and D time to time change how.Wherein, the time increases from left to right, and the line segment of big width is represented bigger distortion.Fig. 6 during with reference to t=0 (i) retrieves the initial coefficients at substation A storage, and the data that receive from substation A are handled.Originally, these data have high distortion levels, but along with the increase of time, this optimization Algorithm program is improved coefficient, make this backoff algorithm reduce aspect the distortion more effectively, and make the distortion from the signal of substation A diminish.When t=2, from the data termination of A, and by the mapping of substation A being stored the improvement coefficient at this substation.Adjust back initial coefficients then at substation B storage, and to this backoff algorithm of data run from substation B, to reduce distortion.Substation B is than the more close central station of other substations (perhaps have the narrower light source of bandwidth or have the light source of " higher standard " with being equal to), makes that the data from B have the distortion levels lower than the primary data that receives from other substations.When t=3, from the data termination of substation B, the improvement coefficient that storage is calculated at this substation, and readjustment is at the new coefficient of substation C.When at t=5, stopping thereby changing into reception, carry out similar step from the data of substation D from the data of substation C.When t=7, at substation A retrieve store when the t=2 before the improvement coefficient.Because therefore the improvement coefficient before having used can be handled processed data more efficiently, it is lower with the distortion levels that obtains to make distortion levels when adopting initial coefficients.In Fig. 6 (i), the narrow linewidth from the data of substation A during by t=7 is represented this situation.In Fig. 6 (i), before the standard operation of carrying out the normal consistency processing, continuously each ONU is trained.This training of backoff algorithm or optimizing process can be handled together with range finding and be embodied as the part of PON start-up period, or be embodied as that the independent compensation of PON after finishing handled or " equalization " startup stage.In case backoff algorithm is with coefficient correlation " locking " and be defined as specific tolerance at a substation, just with these coefficient storage in memory 52 (or local RAM521), in standard operation, to use.In order to simplify, illustration common sequential scheduling.Yet in response to the requirement from substation, the PON scheduling is handled may be more complicated.
From Fig. 6, can understand, adaptive equalization can be (distribute) that disperses, feasible coefficient at a substation carries out the first time to be improved, when the data that next time receive from this substation, keep this through improved coefficient and carry out for the second time, further improve, wherein in middle the period between this improves for the first time and for the second time reception from the data of another substation.
Fig. 6 (iv) illustration alternative " cold " initialization procedure, wherein between a plurality of cells, share (split) each the initial optimization in a plurality of substations handled, and use local RAM to store coefficient correlation again.According to the similar mode of process shown in Fig. 6 (i), Fig. 6 process (iv) may further comprise the steps: (a) coefficient of retrieval substation A; (b) receive data from substation A; (c) use the data that receive from substation A that the coefficient of substation A is improved, and storage is through improved coefficient; (d) substation B, C and D repeating step (a) are arrived (c); And (e) retrieve substation A through improving coefficient.Yet, Fig. 6 (iv) in, since this backoff algorithm stored substation A through improving coefficient, therefore can retrieve these coefficients, make when when at t=4, receiving other data from substation A, this backoff algorithm (its optimizer particularly) can proceed to train and further improve the value of coefficient.In this way, this backoff algorithm can continue from the point that it had before stopped when the t=1.
After during Fig. 6 (i) and 6 (iv) illustrated training or " initialization ", compensating module 44 (backoff algorithm particularly) its signal readjustment this moment is just in the phase relation array of processed ONU.When new ONU is connected to PON, can so that this new ONU by carrying out initialization as the mode of the individual pulse string of Fig. 6 shown in (ii).Yet, in order to minimize influence for other users, preferably, to carrying out this initialization as the (iii) illustrative a plurality of cells of Fig. 6.Wherein, this dispatching algorithm is controlled TDMA and dispersion compensation optimization on a plurality of cells.Yet, storage to weight coefficient makes backoff algorithm can use its previous coefficient of preserving to restart, thereby make it can continue to converge to optimal response, rather than start anew once more that (it is possible and more efficient sending a plurality of cells in same ranging window together, but same basic principle also is suitable for, referring to Figure 12 a and Figure 12 b, Figure 12 a shows the initial phase that sends a plurality of cells in ranging window, and Figure 12 b shows the example that increases the situation of new substation in network).
Be appreciated that and can realize range finding and scheduling by the bit error rate (BER) that is low to moderate 10-4, therefore can be before finishing compensation operation dispatching, thereby can be in the backoff algorithm running application schedules.
In order to make optimizer can more easily improve the value of coefficient (and can select the cranking system array, see below literary composition), preferably, substation sends given data to central station.For example in advance this given data (as PRBS pseudo-random bit sequence) is incorporated in the memory of phase Sub Ammunition Supply Poi in the fabrication stage, make optimizer can carry out feedback or recurrence processing, thus processed data and this given data are compared, and generate convergency value according to the difference between processed data and this given data.In case this convergency value reaches threshold value, just think that coefficient is suitable, and central station keeps the value of the coefficient that is calculated when this convergency value reaches this threshold value.When this situation occurring, central station just prepares to receive data on flows (that is, receiving unknown data from substation).
ISI distortion on the link might time to time change.(may cause the example of this variation to comprise: the transmitter deterioration among the ONU, modal noise, PMD fluctuate, have added new optical fiber and used different transmitters).Therefore, the expectation PON termly (or constantly) weight coefficient of each ONU is optimized again.This can be by realizing according to send special training (non-flow transports) cell termly with physical layer operation, the management wherein used and the similar mode of mode of safeguarding (PLOAM) cell in ATMPON (BPON).Alternatively, can use traffic transport cell itself these coefficients are carried out fine tuning.
For the startup value of coefficient is provided, central station has been stored a plurality of possible startups (equalizer) coefficient sets S1, S2 ..., sj.Can carry out precomputation to them, make each group for example all be suitable for the difference amount that causes owing to different fiber lengths or dissimilar Signal Degrades (ISI).Thus, transport start-up period at non-flow, central station with these not each on the same group (sequentially or concurrently) be applied to into office data, rather than the coefficient sets of algorithm is carried out " training " according to their default value.This Incoming data sequence will be known (for example fabrication stage can with this storage at substation and central station place).So central station only needs by comparing given data and going into the coefficient sets that office data selects to provide optimum performance or minimal error (that is the highest signal of quality).In this way, the cranking system array is tested respectively, and select the coefficient sets of optimum (preferably).Central station can use this selected coefficient sets as the coefficient sets that is used for fixing (non-self-adapting) equalization, or as when using adaptive algorithm, treating improved initial coefficients group.
The flow chart illustration of Figure 13 start-up period.Wherein, at step 0 place, receive given data from new substation.Duplicate this data, and compensate by compensation program (for example following horizontal tap (tap)) each copy to this given data.Use different coefficient sets (or vector) concurrently each copy of this given data to be handled.At step 2 place, the signal quality that uses the data after different coefficient sets are handled is assessed.For example, can be by determining to have received and the high value and the degree of correlation between the low value of " known " data are carried out this assessment (if these data are digital forms) before.At step 3 place, selection and storage provide the coefficient sets of peak signal quality, if equalization (compensation) process is adaptive, then arrival data instant are subsequently used this coefficient sets.Otherwise,, then storing this coefficient sets so that before step 7 place uses, this coefficient sets is improved or " adjusting " after a while in step 4 and 5 places if equalization is adaptive.
The flow chart illustration of Figure 14 the algorithm that wherein different coefficient sets is compared by serial mode.Wherein, at step 1 place, each cranking system array (vector) is applied to each in a plurality of copies of known arrival data.Signal quality to the data through so compensating is assessed, in case all coefficient sets are all tested, just according to retrieving the coefficient sets that obtains highest signal quality, for future use with the similar mode of Figure 13 (it shows by parallel processing and selects coefficient).
Figure 15 (i) and (ii) illustration amount distortion differentiation in time under the situation of from a plurality of one of may organize, having selected the cranking system array.(in the stage (a), the altimeter of bar shaped is shown in and uses different startup groups to compensate residual distortion amount afterwards, and in the stage (b), the height of bar shaped represent to utilize using adaptive algorithm to carry out selected coefficient sets under the improved situation residual distortion amount after compensating).In order to simplify, among Figure 15 (i) the serial realization example has been shown, represented how to use successively in the different coefficient sets each, select best group then, and use adaptive technique further it to be optimized subsequently.
Figure 15 (ii) illustration the situation of PON of wherein a plurality of ONU shared bandwidths, show wherein between initial " cold " starting period of before the standard operation of operation dispatching each ONU being trained serially.(this training process handle can be waited together with range finding be embodied as the part of PON start-up period, or be embodied as independent " equalization " of PON start-up period after finishing startup stage).In case this equalization algorithm " locking " has also been determined the coefficient correlation of ONU, just with these coefficient storage in local RAM, in standard operation, to use.(in order to simplify, illustration common sequential scheduling handle.In response to the requirement from ONU, the PON dispatching algorithm can be more intelligent).The diagram that Figure 15 provides shows the serial of using different coefficient sets to carry out by equalizer and attempts.Handle if can carry out multithreading, then can test a plurality of coefficient sets simultaneously, shortened and be used for determining the required time of optimal set.
Use a plurality of startup groups should " through improved initialization procedure " be strong enhancing because it has overcome misgivings to the length of training sequence-no longer need long sequence, because this device only need be selected the optimum system array rather than optimize fully.Utilize the definition coefficient sets of fair amount, also overcome at network and must transmit the potential engineering restriction that may require under other situation of a certain lowest performance level.
Another advantage that a plurality of cranking system arrays are tested is: even to such an extent as to when the distortion factor can't be found range and/or dispatch too greatly, also can carry out this test.Particularly, can under the situation of the sign of not knowing to initiate substation, test the startup group.If distortion (for example bit error rate) to such an extent as to the too high central station of initial level can not correctly read the arrival data, this situation will appear.In case found suitable coefficient sets, just can use this to organize the arrival data that read subsequently, if contain described sign in feasible these arrival data, then central station can obtain this sign.
An example of backoff algorithm is called as transversal filter and handles.In general, the operation of this processing is included in a plurality of somes time location or " tap " of Incoming data flow (that is, along) and locates to sample.Tap can be according to separating than a little interval (for example each of bit period half or 1/4th) of bit in the Incoming data flow.In addition, the sampling interval needs not to be regular, and can be irregular, takes into account so that for example will arrive the complexity of data.Preferably, tap is positioned at continuously or the bit locations of " adjacent ".So,, weighted sample and this target bits of the adjacent bit of this target bits can be mixed for each target bits to be compensated.Particularly, usually with weighting adjacent bit and target bits addition, or from target bits, deduct the weighting adjacent bit.(because the overlapping between distortion or the bit, the value of bit no longer is 0 or 1, but can be the value in the successive range).
Preferably, adjacent bit will comprise at least one immediate neighbor of target bits, for example handle for 3 tap filters, and target bits and two (being (trailing) that trails in this example) neighbours in data flow thereof are sampled.For 5 bit taps, will be comprised among the neighbours that sample that ensuing 4 of target bits trail neighbours recently, or the like.For the sampling bit is weighted, a weighting function is applied to each corresponding bits, at this weighting function of each bit by with corresponding group of data substation to be corrected in coefficient in corresponding one control.Preferably, this weighting function is simple factorization function (a for example multiplication).
In order to make filter process carry out work, compensating module will for example be stored in the shift register 522 in the memory 52 to the respective memory locations of bit storage in memory 52 of the continuous arrival of determined number.At work, (i) each continuous data bit is weighted by the coefficient that is associated with memory location that this data bit is positioned in fact; (ii) preserve value through weighting; And (iii) then these data are shifted, make that the data in the bit time slots (promptly having arrived next time slot of time slot) that data in the memory location are directly trailed are replaced.Repeating step (i), (ii) and (iii) in order, the result has produced a value for each circulation, and this value is the combination of the data from the sample survey at each memory location place.Particularly, for each target data bit, this combination all comprises this target data bit and the weighting bit of trailing this target bits in specifying neighbours.
More particularly, the data that receive at the central station place can be represented by R (x), and at each bit locations value R1, R2, R3 ..., Rx-1, Rx, Rx+1 etc., Rx is a target bits to be compensated.Wherein, the position of mark " x " expression bit in bit time slots stream." through compensation " data/" output " E (x) is provided by the bit sum of respectively sampling of having carried out weighting by coefficient, that is, is given by the following formula:
E ( x ) = Σ i = 0 n - 1 c i R ( x - i )
Therefore, for " 3 tap " filter (wherein n=3), provide through offset data by following formula:
E (x)=c 0R x+ c 1R X-1+ c 2R X-2It also is useful using vector symbol E=C.R, and wherein C comprises coefficient (c 0, c 1, c 2..., c N-1), and R is a data from the sample survey R x R x - 1 R x - 2 . . . R x - n .
Fig. 7 shows the table that the key step of backoff algorithm is summarized.Coefficient c0, c1 and c2 (they are static in this example) correspond respectively to memory location M1, M2 and M3.The row that is marked as M1, M2 and M3 shows for each step, and which data bit R1-Rx is written among the M1-M3 of corresponding memory position.(right-hand side) row show the offset that obtains at the equalization data at last: for example provided the equalization data at R3 bit time slots position at stage 3c place.
Below provided a kind of description of possible backoff algorithm, this backoff algorithm is adaptive, makes this algorithm to carry out and optimizes or training program, improves with the value to the coefficient that is used for the distorted signals from given substation is compensated.Still by the above formula that provides " through compensation " data/" output " E (x) are described, although allow index variation now and coefficient is regulated improved filter " filtering " to be provided or to reduce distortion.Formula is made amendment with the time-dependent characteristic of reflection coefficient, has provided:
E ( x ) = Σ i = 0 n - 1 c i ( x ) R ( x - i )
Therefore, for " 3 tap " filter (wherein n=3), provide through offset data by following formula:
E(x)＝c 0(x).R x+c 1(x).R x-1+c 2(x).R x-2
When initial, all memory locations all are made as zero.Coefficient is set to their initial value (this is their " expectation "/mean value normally, to help to shorten convergence time).These coefficients are regulated will be referred to the algorithm that certain expression to signal quality responds.If some data is known, then this algorithm can be the error function of only difference between " through compensation " data and this given data being measured.In this case, this optimization manages to reduce this error.Here, use known training sequence so that can directly measure this error.Yet (, can attempt " blind (blind) " is optimized by using such as another signal quality metrics (metric) of " eye is opened (eye opening) ").
Imagination by K (x) provide known, send data.Can be write it as K1, K2, K3 ..., Kx-1, Kx, Kx+1... etc.The data that received are still provided by R (x), and value R1, R2, and R3 ..., Rx-1, Rx, Rx+1... etc.Through data E (x) the value E1 of compensation, E2, E3 ..., Ex-1, Ex, Ex+1.Error e between offset data and given data (x) is exactly K (x)-E (x).This adaptive algorithm will be regulated coefficient according to the mode that makes this error minimize.
Can use various optimisation technique: example is to reduce this error (if form of known this function by coefficient being carried out the increment that trickle change and selection reduce this (all sides) error, then can use this function derivative to carry out the estimation of more meaningful (informed), to revise these coefficients).Use vector symbol, if E=C.R then can be write as revised estimation E '=(C+ δ C) .R, make e ' (=K-E ')＜e (=K-E).New coefficient is provided by C '=C+ δ C.
As summary, carry out following steps (if in several leading circulation data transformation then can omited steps VI and VII) to (using given data) described processing:
When I. initial, all data storage positions all are made as zero, and coefficient is got initial value
II. data are shifted
III. the signal at position x place is sampled, it is saved in first Data Position
IV. calculate data E through compensation
V. adjust back given data and error of calculation e
VI. calculate " gradient ", that is, determine coefficient change so that error reduce or minimize
VII. be with coefficient adjustment as determined new value among the VI
VIII. forward Step II to
The function that can be used as the offset delay value that central controller calculated at each substation calculates initial value, perhaps can use another delay relevant parameter of the transit time of expression from the substation to the central station, because this transit time is associated with fiber path length usually, fiber path length and then represented possible distortion level.Fig. 8 a and 8b show the table of step related in the expression adaptive compensation algorithm.But Fig. 8 a shows going together mutually of this table different lines with 8b.
Below be to the description of explicitly permission from the algorithm of the data demand of different substations different filter factor c1, c2, c3 etc.Formula by the above self adaptation example that provides is described " through compensation " data/" output " E (x), although now all have independent coefficient sets at each ONU.Filter works by identical mode, jumps between the data set of different ONU but be to use from the knowledge of the sending order of this dispatching algorithm.Write as the processing of step one by one:
When I. initial, all data storage positions all are made as zero, and coefficient is got the initial value normal running
II. new cell one uses self scheduled algorithm's knowledge to determine which ONU is sending
III. retrieve the coefficient of ONU
Data are carried out equalization
IV. data are shifted
V. signal is sampled, it is saved in first Data Position
VI. calculate data E through compensation
If data are known, then coefficient is optimized (otherwise forwarding step X to)
VII. adjust back given data and error of calculation e
VIII. calculate " gradient ", that is, the change of determining coefficient is so that error minimize
IX. be with coefficient adjustment as determined new value among the VIII
Loop ends
If X. arrive the cell end, packing coefficient and forward Step II to then, otherwise forward step IV to
Fig. 9 a, 9b with the 9c illustration in data the table (but 9a, 9b show going together mutually of this table different lines with 9c) under the known and situation that coefficient is regulated.If data are not known, then skips steps VII fixes to IX and retention coefficient.
Modeling process shown discussed 10 -3To 10 -4BER realize down the initialized feasibility of PON.Following table 1 shows under this BER the failure-free probability that receives the data that are used to start (startup, range finding etc.), although performance is unacceptable for customer traffic.
Log(BER) Obtain the probability of n continuously no error code word joint
?n＝1 ?n＝2 ?n＝3 ?n＝4 ?n＝5 ?n＝10 ?n＝20 ?n＝30 ?n＝40 ?n＝43
1 ?0.4305 ?0.1853 ?0.0798 ?0.0343 ?0.0148 ?0.0002 ?5E-08 ?1E-11 ?2E-15 ?2E-16
2 ?0.9227 ?0.8515 ?0.7857 ?0.725 ?0.669 ?0.4475 ?0.2003 ?0.0896 ?0.0401 ?0.0315
3 ?0.992 ?0.9841 ?0.9763 ?0.9685 ?0.9608 ?0.9231 ?0.8521 ?0.7865 ?0.726 ?0.7088
4 ?0.9992 ?0.9984 ?0.9976 ?0.9968 ?0.996 ?0.992 ?0.9841 ?0.9763 ?0.9685 ?0.9662
5 ?0.9999 ?0.9998 ?0.9998 ?0.9997 ?0.9996 ?0.9992 ?0.9984 ?0.9976 ?0.9968 ?0.9966
6 ?1 ?1 ?1 ?1 ?1 ?0.9999 ?0.9998 ?0.9998 ?0.9997 ?0.9997
Table 1
Last table shows to have good probability from communicate required limited quantity byte reception data in the initial activation stage before carrying out the equalization processing.If fault, then ONU can send simply again.Following table (table 2) shows the probability of success of attempting increasing afterwards 5 times.
Log(BER) After 5 times are attempted, obtain the probability of n continuously no error code word joint
?n＝1 ?n＝2 ?n＝3 ?n＝4 ?n＝5 ?n＝10 ?n＝20 ?n＝30 ?n＝40 ?n＝43
1 ?0.9401 ?0.6411 ?0.3401 ?0.1603 ?0.0718 ?0.0011 ?2E-07 ?5E-11 ?1E-14 ?0
2 ?1 ?0.9999 ?0.9995 ?0.9984 ?0.996 ?0.9485 ?0.6729 ?0.3747 ?0.1851 ?0.1479
3 ?1 ?1 ?1 ?1 ?1 ?1 ?0.9999 ?0.9996 ?0.9985 ?0.9979
4 ?1 ?1 ?1 ?1 ?1 ?1 ?1 ?1 ?1 ?1
Table 2
Figure 11 illustration the key step in the adaptive algorithm of Fig. 9 a, 9b and 9c.Figure 11 shows the alternative of adaptive compensation algorithm and represents.Wherein, when initial, the data from the input stage 40 of central station are carried out " sampling " with extraction bit time slots value, and coefficient of utilization c0 is weighted to it.Before adding to sample to next bit time slots in power level 902 places and with coefficient c1 it being weighted, using the function with a bit delay by conversion stage 901 (is conversion Z here -1).Used another conversion stage of the delay with further bit delay before data are sampled, by another coefficient it is weighted then, for an other bit delay, the rest may be inferred.At sum unit 903 places the sampling through weighting is sued for peace then.Here, conversion stage 901, to add power level 902 and sum unit 903 are the software units that are implemented in (for example being arranged in the processor device 50 of central station and storage arrangement 52) processor and memory.So coefficient sets can be changed over the next one from an ONU.
Can use other equalization (compensation) algorithm family, for example: linear equalization; Decision feedback equalizationization (and feedforward); Non-linear equalization; Maximum likelihood sequence detection (MLSD) is also referred to as Maximum Likelihood Detection (MLD); Possible " blind " technology; And constant modulus algorithm (CMA).
With reference to following steps (wherein central station is called OLT, substation is called ONU), with the mode embodiment that the present invention may be better understood of example.
1.OLT all ONU are broadcasted.After ONU powered on, ONU intercepted the downstream flow from OLT, and ONU can obtain bit and frame synchronization (it can obtain position and frame synchronization according to the data of minute frame) according to this operation
2.OLT all ONU are broadcasted.(a plurality of) new ONU intercepts and receives all the broadcasting cells from OLT.ONU extracts network parameter (form of initial transmission power rank and range finding correlation delay, frame, treat the preamble that uses the upstream etc., when send the details of registration and distance measurement request etc.).OLT is regularly broadcasting these parameters in the define byte downstream, makes long enough time of just intercepting that ONU will do to extract relevant information.For example, OLT can send byte in the define byte of head, occurred ranging window with the byte XXX that is illustrated in this frame between the YYY.(alternatively, these bytes can be represented the end TTT period afterwards of frame ZZZ, perhaps the TTT period after OLT has sent given byte sequence).
3. be appreciated that at any single ONU place,, may have experienced identical distortion (ISI) from the signal of OLT for its tdma slot.Each frame or cell all have the known array byte that can be used as the input data of backoff algorithm.Alternatively, can send the cell of the given data that is exclusively used in compensation according to the interval (for example per 128 frames) of rule.Which kind of mode no matter, (a plurality of) ONU can use this go into office data to it/their backoff algorithm trains, and their coefficient is optimized.Head end (head end) comprises the laser of higher standard, so it can use the ISI compensation on downstream direction.Yet, if desired, also can keep the selection of using the ISI compensation.This step also is convenient to execution in step 7, and step 7 helps to quicken the convergence of this backoff algorithm at updrift side.
4.ONU its setting is adjusted to those settings that receive from OLT, and [in (1), make its own and OLT synchronously after], determine that wherein it can attempt registering and the available window of range finding etc. (some PON agreement does not require and finds range and register, but this is more generally to require).
5. in log-in window, ONU will send request to OLT and register to use in the pre-determined factor group one or more.Diverse ways can be arranged: (A) ONU is applied to data simultaneously concurrently with each coefficient sets.Some result can be destroyed fully by ISI, but some result will provide good data.ONU selects to provide the group of optimum performance.Can accept then and fixing selected coefficient, perhaps used as the initial input parameter, further to optimize by the adaptive equalization algorithm; (B) ONU can be in same starting/in the range finding cycle each coefficient sets sequentially is applied to data.Equally, some the possibility of result is destroyed, but ONU only selects to provide the group of optimum performance.Can accept and fix these coefficients then, perhaps used as the initial input parameter, so that further optimize by the adaptive equalization algorithm; (C) if ranging window is too short, then ONU can alternatively send a plurality of requests on continuous ranging window.OLT will use different coefficient sets in each log-in window, find up to it to provide performance enough till the group of initiating between OLT and the ONU to communicate by letter.OLT can wait for that till it had used each coefficient sets, to select best coefficient sets, perhaps after having set up communication, the method for " intelligence " of can using was more determined best group to be used.
As optional improvement, if in step 1 to 3 when intercepting from the downstream data of OLT, ONU has determined then this information upstream to be sent to OLT to the measuring or coefficient that the equalization algorithm the is required of ISI on the link.In case one in the above-mentioned several different methods has been found the one group of coefficient that is enough to extract these data, then the coefficient that this information can be used for predicting according to ONU is selected the optimum system array.
After having selected to be used for the suitable coefficient of equalization algorithm, OLT can be from Incoming extracting data information.These data have comprised the request from ONU.OLT handles (using suitable security measures if necessary) to this request, and broadcasts suitable response.Described (a plurality of) ONU intercepts the response in the suitable downstream byte.If registration failure, then ONU can repeat to attempt in log-in window after a while.
6. range finding.Set up ranging window, make other ONU not send, so irrelevant disturbing, and do not interrupted by range finding/activation.In some agreement such as GPON, the ONU that in the range finding processing procedure that ONU carries out one by one, stops to work.For alternative protocol, the ONU that is working is stopped.In " range finding " window, ONU sends message in the known time with respect to the startup of ranging window.OLT receives this message, and determines its time of advent, according to this time of advent, utilizes knowledge to the delay at ONU place (that is, when sending response with respect to the startup of ranging window) can calculate round-trip delay, and so can calculated transmission delay.Use this transmission delay to determine to need to increase great additional delay to guarantee that its upstream data is synchronous then at the ONU place.Then this information is sent to ONU downstream.(for other downstream datas, together with making ONU can determine that the suitable identifier of its data comes it is broadcasted).
7. to the optimization of equalization coefficient.Suppose in step 5, to have determined to compensating the good estimated initial of (equalization) coefficient.This may be enough in simple the realization, does not therefore need these coefficients are carried out any further adjusting, in this case, can carry out next step.Yet, if desired, also can carry out further and optimize the equalization coefficient.There are many selections equally.
A kind of method is broadcasting one message, so that every other ONU stops, and makes new ONU can send the training sequence of sufficiently long known byte to OLT.In case OLT optimizes penalty coefficient at this new ONU, has perhaps reached the maximum period, then store these penalty coefficients for this ONU.OLT then can broadcast another downstream message that the expression compensation finishes, and makes it possible to another ONU is registered, finds range or compensates, and perhaps makes it possible to turn back to normal running.
A kind of alternative methods is as the part of range finding message or defines the training sequence of the regular length that can send in " compensation " window independently.
Another kind method is to use extended message in ranging window, because this has been avoided owing to additional window makes agreement complicated more.This is possible, because comparatively speaking, has a lot of up durations in ranging window.The size of necessary correct specified window, to handle these long message, still required overhead should be very little.If desired, can use collision detection to help handle the situation that different ONU use same ranging window: promptly, can use collision detection multiple access (CDMA) to guarantee once to have only an ONU to attempt carrying out equalization.Under situation about clashing, ONU retreats (back off) random quantity, and ranging window is monitored, and makes once to have only an ONU to attempt finding range and equalization.Because register this ONU and find range this moment, so can control it, therefore can dispatch it now, thereby can preserve penalty constant for this ONU by the PON dispatching algorithm, and if desired, can make compensation optimizing expand to a plurality of range findings/optimization window.
After setting the optimization of quantity,,, just can think and finish initial optimization in case perhaps the increment of coefficient or error change drop to below the specific size in case error drops to below the specific size.In course of normal operation, can realize further optimization, referring to following steps 9.
8. in case OLT has carried out optimizing to the compensation of ONU and finished registration and range finding, it just can represent that it can register another ONU/find range and the renewal of equalization to other ONU broadcasting.(this step is optionally, especially when training sequence has regular length and be positioned at definite (define) position with respect to downstream frame).If there is no " conflict " (one of them above ONU sends simultaneously) then can the more than one optimization of single treatment.
9. normal running.Network is found range, make all upstream messages all synchronous together, thereby they can arrive their trams in frame respectively.Also ONU is compensated.There are multiple choices in operation for subsequently.In the best group of having selected pre-determined factor, and there are many selections in (as mentioned above) for " normal running " also alternatively after mounting phase has carried out further optimization to them:
A) coefficient can be maintained fixed for this ONU, makes that OLT uses identical coefficient sets when ONU sends.
B) OLT can be by once in a while to from a plurality of predetermined group (whole in them, perhaps only be in the predetermined group of the most approaching current predetermined group those aspect the ISI compensation) signal quality compare, attempt once in a while the pre-determined factor group that it uses is optimized again.
C) OLT can attempt once in a while coefficient itself is optimized again according to previous realization.
For B) and C), OLT can use the known byte in the head of known " training " cell of normal data cell, PLOAM cell or other special uses to carry out this optimization.B) and the optimization again C) can regularly carry out, for example every x minute, every day etc., perhaps it can be that tolerance when signal quality drops to the processing of being aroused (prompt) under the situation that is lower than certain threshold level.(this threshold value can be an absolute value, or after optimizing with respect to the value of initial performance).
Current this generation PON carries out work with 622Mbit/s usually, has the maximum magnitude of 32 shunt (that is every OLT32 ONU) and 20km usually.Yet the PON of a following generation may carry out work with higher bit rate and on longer distance, to walk around outer core transmission (outer core transmission) equipment.The bit rate of the 10Gbit/s that this may relate to (but being not limited to) on the fiber distance of～100km.Even in current system, think that also distortion or other inter symbol interference (ISI) (it may be owing to causing such as the factor along CHROMATIC DISPERSION IN FIBER OPTICS (CD) or polarization mode dispersion (PMD)) are undesirable, and the light source that has a wideer live width under high bit rate more or for substation wherein (promptly, " low specification " light source) system, this distortion or ISI may be more undesirable.
Can see, among the above embodiment some has utilized the knowledge of the PON multiple access protocol that the order of each OLT of tdma slot sends at least.As the part that scheduling is handled, OLT has known in advance next message source to be received is from which ONU.Equalizer (compensation process) can utilize this knowledge to switch to the weight coefficient of next ONU fast between suitable weight coefficient, to compensate the ISI of next cell.This means that adjacent cell may have very different ISI amount,, but can adjust back last group coefficient apace and therefore can in the equalization algorithm, keep suitable value because this algorithm needn't be for the optimization process that starts anew of each cell.Thus, a plurality of (for example 10) that OLT will have an expression different I SI amount are the equalization coefficient sets fixedly, rather than must carry out " training " fully to the adaptive equalization algorithm, to be optimized at the coefficient of mounting phase to this algorithm.(this for example can corresponding to different fiber lengths).At mounting phase, this device only is applied to each parameter group in these different parameters groups (sequentially or concurrently) into office data, rather than begins the coefficient of this algorithm is carried out " training " from their default value.Because data sequence is known, so this device only need select to provide the coefficient sets of optimum performance/minimal error.It can use this fixed coefficient group as the coefficient sets that is used for fixing (non-self-adapting) equalization subsequently, perhaps as treat improved initial coefficients group when using adaptive algorithm.
The useful consequence that adopts electronic equalizingization is to have parameter or the coefficient that is used by this equalization algorithm in the memory, and can easily these parameters or coefficient be copied to output, in this output place, they can provide the monitoring function from the signal quality of multiple ONU.This is helping to provide to will being useful instrument for Virtual network operator aspect the monitoring of PON.In this feeds of data can being used to the secondary of understanding PON topological structure (it can make the data from different ONU be associated), fault among the PON and deterioration are positioned with help.
In the foregoing description one or more will be convenient to existing PON (for example, according to the PON that carries out work (at least before the present invention) such as the standard of ITU G983 standard, and/or the PON that carries out work with high bit rate) is operated.This is that distorted characteristic realizes in the data that change on the short time yardstick (for example because the path of these data on this short time yardstick variation has taken place) by allowing to be applied to compensation program wherein.The PON of a current generation is tending towards utilizing high quality optical assembly (for example external modulation and distributed Feedback (" DFB ") laser) to help overcome the influence of the Signal Degrade that causes owing to ISI.As mentioned above, the more cheap assembly such as Fabry-Perot (Fabry Perot) laser is used in expectation, but the performance of their relative mistakes has often hindered their application.The foregoing description makes it possible to use the assembly of low specification, makes the operator that manages to dispose access network can save considerable cost.Potential saving is significant, and can change the economy of PON configuration, makes it become the feasible program of access network.This advantage may have than the high bit rate of being discussed, the bigger potential application of long distance applications.
Further consider to comprise:
Preferably, receiver and equalization algorithm can be handled the burst mode flow.
Can use other equalization (compensation) algorithm family, for example: linear equalization;
Decision feedback equalizationization (and feed-forward); Non-linear equalization; Maximum likelihood sequence detection (MLSD) is also referred to as Maximum Likelihood Detection (MLD); Possible " blind " technology; And constant modulus algorithm (CMA).
Above embodiment can be used for other ISI sources: relatively poor transmitter extinction ratio; (time change) PMD; Mode segmentation fluctuation (mode partitionfluctuation) with the chromatic dispersion combination.
Using the possible alternative arrangement of RAM will be to use multithreading in parallel processor.
In another embodiment, can monitor coefficient (parameter), to monitor the performance of one or more transmission paths.

Claims (23)

1, a kind of central station that is used for receiving data from a plurality of substations, described central station is configured to carry out in use and is used for compensation process that the deterioration from the data of described a plurality of substations is compensated, described compensation process has at least one tunable characteristic by parameter group control, wherein, described compensation process may further comprise the steps: (i) use different start-up parameter groups that the data from substation are compensated; (ii) the quality of using the described data after described different start-up parameter group compensates is measured; And, select a start-up parameter group that the data from the follow-up arrival of described substation are compensated (iii) according to measured quality.
2, central station according to claim 1, wherein, described central station is configured to store a parameter group at each substation.
3, central station according to claim 2, wherein, arriving according to step (i) (iii) is that each substation is selected institute's stored parameters group.
4, according to any described central station in the aforementioned claim, wherein, execution in step (i) is to (iii) in response to the reception of the data of the substation that connects making a fresh start.
5,, wherein,, come from other start-up parameter groups, to select a parameter group by the quality of using described different start-up parameter group to obtain is compared according to any described central station in the aforementioned claim.
6, according to any described central station in the aforementioned claim, wherein, use and the start-up parameter group is assessed, preferably from the test data of substation, before the described test data from substation arrives, the copy of described test data is stored in described central station.
7, according to any described central station in the aforementioned claim, wherein, described parameter group includes a plurality of parameters.
8, according to any described central station in the aforementioned claim, wherein, the substation that described central station is configured for from the overall optical network receives data.
9, according to any described central station in the aforementioned claim, wherein, described compensation process is carried out in electric territory.
10, according to any described central station in the aforementioned claim, wherein, described compensation process may further comprise the steps: from a plurality of time locations place in the data flow of substation to the described line sampling that flows to; And to each sample execution function corresponding.
11, according to any described central station in the aforementioned claim, wherein, described central station is configured to send dispatch command to substation, and to receive data from this substation, described dispatch command comprises this substation of permission and sends the order that data reach the fixed time.
12, central station according to claim 11, wherein, described central station is configured to send further dispatch command, so that can send this further data under the further data conditions of needs.
13,, wherein, be to come selected according to institute's stored scheduling instruction to going into the parameter group of using when office data is used described compensation process according to claim 11 or the described central station of claim 12.
14, according to any described central station in the claim 11 to 13, wherein, described dispatch command comprises and has identified the identifier that allows to send from it substation of data, and wherein uses the described identifier to retrieve the parameter group that is associated with the substation that is identified.
15, according to any described central station in the aforementioned claim, wherein, described compensation process comprises adaptive algorithm, this adaptive algorithm is constructed such that when the data that receive from substation, use the described data that receive, the value of the parameter that is used for this substation is improved with respect to one group of initial value.
16, central station according to claim 15, wherein, described central station is constructed such that in the source that arrives data when first substation changes over second substation, described central station: (i) storage is at the improvement value from the parameter of the data of described first substation, to be used for the retrieval of back; (ii) retrieval is before at the described second substation stored parameters; (iii) in response to the further data from the expection or the arrival of described first substation, retrieval is at the previous improvement value of the parameter of described first substation; (iv) and preferably the parameter to described first substation is further improved.
17, the method for a kind of operation communication system, this communication system comprises central station and is connected to a plurality of substations of described central station, described central station can be used for carrying out and is used for the compensation process that carries out from the deterioration of the data of described a plurality of substations, described compensation process has at least one tunable characteristic by parameter group control, said method comprising the steps of: (i) use different start-up parameter groups that the data from substation are compensated; (ii) the quality of using the data after described different start-up parameter group compensates is measured; And, select a start-up parameter group that the data from the follow-up arrival of described substation are compensated (iii) according to measured quality.
18, method according to claim 17, wherein, described communication system comprises the optical-fiber network that is used for described central station is connected to described a plurality of substations.
19, according to claim 17 or the described method of claim 18, wherein, described optical-fiber network comprises and being used for from the signal multiplexing of at least two substations branch's contact to the common light carrier wave, described a plurality of substation is formed at a plurality of time intervals and sends data, makes in use data from described a plurality of substations by passive time division multiplexing.
20, according to any described method in the claim 17 to 19, wherein, the storing predetermined data of described central station, and described or each substation is stored corresponding tentation data, and wherein said or each substation sends to described central station with its tentation data, makes the described tentation data that is stored in described central station to be compared with the tentation data that is received.
21, method according to claim 20 wherein, compares tentation data of being stored and the tentation data that is received, so that the quality of using the data after described different start-up parameter group compensates is measured.
22, according to claim 20 or the described method of claim 21, wherein, the described tentation data that is stored in substation comprises at least some total data with the described tentation data that is stored in described central station.
23, a kind of communication system that comprises central station and a plurality of substations, described central station is constructed to carry out and is used for compensation process that the deterioration from the data of described a plurality of substations is compensated, described compensation process has at least one tunable characteristic by parameter group control, wherein, described central station stores at least one parameter group at each substation, and for each substation, described central station uses the parameter group that is associated with this substation that the data from this substation are used backoff algorithm.
CN2005800108034A 2004-03-31 2005-03-30 Communication system Active CN1939020B (en)

Applications Claiming Priority (5)

Application Number Priority Date Filing Date Title
GB0407341A GB0407341D0 (en) 2004-03-31 2004-03-31 A communication system
GB0407341.7 2004-03-31
GB0502257.9 2005-02-03
GB0502257A GB0502257D0 (en) 2005-02-03 2005-02-03 A communication system
PCT/GB2005/001252 WO2005096574A1 (en) 2004-03-31 2005-03-30 A communication system

Publications (2)

Publication Number Publication Date
CN1939020A true CN1939020A (en) 2007-03-28
CN1939020B CN1939020B (en) 2012-07-04

Family

ID=32247623

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2005800108034A Active CN1939020B (en) 2004-03-31 2005-03-30 Communication system

Country Status (2)

Country Link
CN (1) CN1939020B (en)
GB (1) GB0407341D0 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101989879A (en) * 2009-08-05 2011-03-23 华为技术有限公司 Electronic dispersion compensation method, device and system in burst mode
CN102064928A (en) * 2010-12-10 2011-05-18 福建星网锐捷网络有限公司 Method and device for obtaining signal compensating parameters and network equipment
CN109075861A (en) * 2016-06-14 2018-12-21 华为技术有限公司 Upstream data equalization methods, device and system

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TW282601B (en) * 1995-01-24 1996-08-01 Ibm
AU9213398A (en) * 1997-12-12 1999-07-05 Thomson Consumer Electronics, Inc Tdma wireless telephone system with independently tracked demodulation parameters
JP4257058B2 (en) * 1999-08-23 2009-04-22 富士通株式会社 WIRELESS COMMUNICATION SYSTEM, RECEIVING FIELD INTENSITY COMPENSATION METHOD IN WIRELESS COMMUNICATION SYSTEM, BASE STATION AND TERMINAL STATION FOR WIRELESS COMMUNICATION SYSTEM

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101989879A (en) * 2009-08-05 2011-03-23 华为技术有限公司 Electronic dispersion compensation method, device and system in burst mode
CN101989879B (en) * 2009-08-05 2013-09-11 华为技术有限公司 Electronic dispersion compensation method, device and system in burst mode
CN102064928A (en) * 2010-12-10 2011-05-18 福建星网锐捷网络有限公司 Method and device for obtaining signal compensating parameters and network equipment
CN109075861A (en) * 2016-06-14 2018-12-21 华为技术有限公司 Upstream data equalization methods, device and system
US10506313B2 (en) 2016-06-14 2019-12-10 Huawei Technologies Co., Ltd. Upstream data equalization method, apparatus, and system cross-reference to related applications
CN109075861B (en) * 2016-06-14 2020-04-28 华为技术有限公司 Uplink data balancing method, device and system

Also Published As

Publication number Publication date
GB0407341D0 (en) 2004-05-05
CN1939020B (en) 2012-07-04

Similar Documents

Publication Publication Date Title
Arik et al. MIMO signal processing for mode-division multiplexing: An overview of channel models and signal processing architectures
CN103155450B (en) Method and apparatus for determining the asymmetry in communication network
JP5086334B2 (en) Method and apparatus for incorporating an outstation into an optical network
US9025949B2 (en) Equalization delay agnostic protection switching in protected passive optical networks
CN101160757B (en) Common electronic dispersion compensation device for multiple optical communication channels
EP2997677B1 (en) Optical network unit self-calibration in multi-wavelength passive optical network
US20100310252A1 (en) Optical communication
US20190089463A1 (en) Using multi-level modulated signals in passive optical networks
CN101364844A (en) Method, system and apparatus realizing remote data transmission in passive optical network
US8600235B2 (en) Network with sourceless clients
NO302500B1 (en) Stations for a digital communications network
US8081877B2 (en) Method of determining an optical distance for chromatic dispersion compensation
KR101093856B1 (en) A communication system
KR102112785B1 (en) Wavelength selection system and method for multi-wavelength passive optical network
CN1939020B (en) Communication system
CN101459472A (en) Method, system and apparatus realizing remote data transmission in passive optical network
JPH01126037A (en) Optical remote communication system
EP1972085B1 (en) Optical communication network and method of routing in the network
Baziana Resource allocation mechanism for efficiency improvement in WDM networks: an analytical comparison study
Horvath et al. Simulations of grant allocation in NG-PON2 networks using OPNET modeler
Lohani et al. Design and simulation of hybrid WDM/TDM passive optical network
Kahn et al. MIMO channel statistics and signal processing in mode-division multiplexing systems
Maeda et al. High-speed QKD system synchronized by automatic phase-alignment mechanism
KR101115251B1 (en) Time Division Multiple Passive Optical Network and its compensation method of routes difference and the recording media storing the program performing the said method
US11916747B1 (en) Systems and methods for automatic discovery of a communication network

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant