WO2010069390A9

WO2010069390A9 - Method and device for data processing in an udwdm network and communication system comprising such device

Info

Publication number: WO2010069390A9
Application number: PCT/EP2008/067992
Authority: WO
Inventors: Erich Gottwald; Sylvia Smolorz
Original assignee: Nokia Siemens Networks Oy
Priority date: 2008-12-19
Filing date: 2008-12-19
Publication date: 2011-05-12
Also published as: WO2010069390A1

Abstract

A method and a device for data processing in an UDWDM network are provided, wherein one of at least two OLTs communicates with at least one ONU, wherein the at least one ONU is connected to a splitter that is connected to each of said at least two OLTs, wherein the at least two OLTs are active, and wherein the at least two OLTs provide protection switching and/or load sharing. Furthermore, a communication system is suggested comprising said device.

Description

Method and device for data processing in an UDWDM network and communication system comprising such device

The invention relates to a method and to a device for data processing in an UDWDM network and to a communication system comprising such a device. A passive optical network (PON) is a promising approach regarding fiber-to-the-home (FTTH) , fiber-to-the-business

(FTTB) and fiber-to-the-curb (FTTC) scenarios, in particular as it overcomes the economic limitations of traditional point-to-point solutions.

The PON has been standardized and it is currently being de^¬ ployed by network service providers worldwide. Conventional PONs distribute downstream traffic from the optical line ter^¬ minal (OLT) to optical network units (ONUs) in a broadcast manner while the ONUs send upstream data packets multiplexed in time to the OLT. Hence, communication among the ONUs needs to be conveyed through the OLT involving electronic process^¬ ing such as buffering and/or scheduling, which results in latency and degrades the throughput of the network.

In fiber-optic communications, wavelength-division multiplex^¬ ing (WDM) is a technology which multiplexes multiple optical carrier signals on a single optical fiber by using different wavelengths (colors) of laser light to carry different sig- nals. This allows for a multiplication in capacity, in addi^¬ tion to enabling bidirectional communications over one strand of fiber.

WDM systems are divided into different wavelength patterns, conventional or coarse and dense WDM. WDM systems provide, e.g., up to 16 channels in the 3rd transmission window (C- band) of silica fibers around 1550 nm. Dense WDM uses the same transmission window but with denser channel spacing. Channel plans vary, but a typical system may use 40 channels at 100 GHz spacing or 80 channels with 50 GHz spacing. Some technologies are capable of 25 GHz spacing. Amplification op^¬ tions enable the extension of the usable wavelengths to the L-band, more or less doubling these numbers.

Optical access networks, e.g., a coherent Ultra-Dense Wave^¬ length Division Multiplex (UDWDM) network, are deemed to be the future data access technology.

Communication networks need to be protected against failures in order to ensure a continuous communication capability. This is usually done by doubling all or parts of the equip^¬ ment as well as the communication lines. In case of failure, the communication is switched to the second equipment and/or line .

In such UDWDM system, passive optical splitters with high splitting factors are required. Equal splitting for n par- ticipants results in a loss amounting to a minimum of

3*log2 (n) (dB) , e.g., a splitter with n=1024 has a loss of at least 30 dB, in practice due to splice and coupler losses this loss amount to 34 dB . Each outgoing fiber is connected to a transit point of a building, a flat or the like. Carriers of the optical network do not exactly know as how the transmission capacities are to be allocated by their subscribers. Hence, in order to meet potential capacity demands of the subscribers, the operators of optical access networks need to oversize the capacity of a subnetwork that is associated to a splitter. Hence, a sig^¬ nificant amount of bandwidth may at least temporarily be un^¬ used . The problem to be solved is to provide for an efficient utilization of bandwidth in an optical network. This problem is solved according to the features of the inde^¬ pendent claims. Further embodiments result from the depending claims . In order to overcome this problem, a method for data process^¬ ing in an UDWDM network is provided

- wherein one of at least two OLTs communicates with at least one ONU,

- wherein the at least one ONU is connected to a split- ter that is connected to each of said at least two

OLTs,

- wherein the at least two OLTs are active, and

- wherein the at least two OLTs provide protection

switching and/or load sharing.

This approach of data processing in particular allows for an efficient data distribution towards the at least one ONU. However, the traffic from the at least one ONU can also be provided or conveyed to a backbone network via said at least two OLTs.

Said communication of the at least two OLTs (in particular one of the at least two OLTs) with at least one ONU may com^¬ prise transmitting data from the OLTs to the at least one ONU and/or vice versa. In particular, data communication between OLT and ONU (and between OLTs) may be bidirectional. The OLTs and the at least one ONU may exchange data via an optical connection or link, e.g., via an optical network. The approach provided allows for a protection scheme to be cost effective. In addition, a notable fast switching time is facilitated for the UDWDM network in particular in a converged metro and/or access domain. It is noted that there may be two OLTs or several OLTs that may at least partially "protect each other" by means of said protection switching. Hence, if one OLT fails, the remaining OLTs are capable of handling the traffic of the failed OLT. This can be achieved, e.g., by providing a sufficient degree of redundancy between the OLTs. For example, each OLT may keep free a certain number of wavelengths for such redundancy purposes, e.g., in case of failure of another OLT.

It is also noted that said failure is just an exemplary rea^¬ son for triggering protection switching. Another reason may be maintenance of the OLT, failure of a link, etc. Furthermore, said load sharing may use a portion of wave^¬ lengths set aside by a particular OLT. Hence, in case of bandwidth shortage at a particular OLT, at least one of the remaining OLTs may provide wavelengths (channels) to carry traffic of this OLT under shortage. Such service can be pro- vided temporarily until further notice or it may be triggered and/or supervised by a timer. This applies to said protection switching as well.

Hence, the term "provide protection switching and/or load sharing" may comprise the fact that at least one OLT provides redundancy functionality for all or a portion of the remain^¬ ing OLTs, in particular by reserving wavelengths (channels) . Such reserved wavelengths can be used for traffic from an^¬ other OLT in case of this other OLT being (temporarily) out of service or in case of bandwidth or channel shortage.

As at least two OLTs can be "always on", i.e. active, the protection switching as well as the load sharing can be executed almost immediately and without significant delay. This is a decisive advantage compared to a backup-OLT that, e.g., in case of failure, needs to be activated first thereby con^¬ suming a significant amount of time before being able to process traffic of the failed OLT. It is further noted that the OLT mentioned may be a UDWDM OLT. As the ONUs are connected to a splitter, the whole path from the splitter to the backbone network even past the OLTs is protected. Hence, in case of a failure in one path, still a considerable amount of bandwidth can be supplied to and/or from the users (i.e. ONUs) .

According to an embodiment, the splitter is connected via disjunctive fibers to each of said at least two OLTs. According to a further embodiment, several ONUs are connected to the splitter or to an additional splitter, wherein the additional splitter is connected to each of said at least two OLTs . It is yet an embodiment that the splitter is a m:n splitter, in particular a passive m:n splitter, wherein m refers to a number of OLTs that can be connected and n refers to a number of ONUs that can be connected. Hence, according to the m OLTs the ONUs are provided with a m redundancy or protection.

According to another embodiment, the splitter is connected to the at least two OLTs via at least two CWDM splitters, one CWDM splitter per each OLT .

In particular, in each path towards an OLT, the splitter is connected to a separate CWDM splitter. Hence, there may be m CWDM splitter for protection purposes.

In an embodiment, each of the at least one ONU selects one of the at least two OLTs.

Hence, the ONU is free to select an OLT. Accordingly, also load sharing may be initiated by the ONU. For example, if an ONU detects a shortage of bandwidth or an increasing error rate (or a broken link) it may switch over to another OLT. This can be achieved without significant delay as the other OLT is already active.

In another embodiment, wavelengths supplied by the at least two OLTs are at least partially interleaved.

Interleaving wavelengths could be understood as wavelengths that are in an alternating sequence associated with the at least two OLTs. For example, if two OLTs (OLT_l and OLT_2) are active, the total range of wavelengths can be assigned to the first and the second OLT in an alternating order, i.e. OLT_l is assigned the first wavelength, OLT_2 occupies the second wavelength, OLT_l obtains the third wavelength and so on .

This further reduces a delay when the ONU switches from one OLT to another OLT: As the ONU ' s scanning for the next free wavelength is with high probability yet successful at an ad^¬ jacent wavelength (if not, the next free channel may be found two wavelengths ahead, etc.), the ONU may quickly swap to the other (new) OLT.

It is noted that each OLT provides a certain amount of wave^¬ lengths, each of which can be used as a logical channel or a logical link between the ONU and this OLT. The wavelength may be also identified by a particular frequency. The wavelength mentioned - as well as the frequency - may refer to a range or a band, i.e., a particular interval, comprising this particular wavelength (or frequency) .

In a next embodiment, said protection switching and/or load sharing is triggered by at least one OLT and/or by the at least one ONU. For example, a first OLT - in case of a recognized failure

(e.g., a broken connection to the core network) - may trigger switching the ONUs assigned to this first OLT to a second OL . Also, each ONU may - independently - recognize, e.g., a broken link, and switch to a different OLT.

It is also an embodiment that at least one OLT tracks wave- lengths of at least one other OLT, in particular to avoid collisions due to drifting laser sources.

In case of OLTs being operative in an interleaved mode (adja^¬ cent wavelengths being provided by alternating OLTs) , each of two spectrally interleaved OLTs may track the frequencies of the other OLT(s) to avoid collisions due to drifting laser sources .

Pursuant to another embodiment, the at least one OLT follows a wavelength drift of at least one other OLT.

According to an embodiment, the at least one OLT tracks a wavelength drift of the at least one other OLT and informs this at least one other OLT to adjust its wavelength.

According to another embodiment, the at least one OLT is op^¬ tically coupled to the at least one other OLT and at the least one other OLT performs an assessment of a wavelength drift based on the optically coupling and adjusts its wave- length based on the assessment.

In yet another embodiment, the at least one ONU is associated with at least one backup-wavelength. Such at least one backup-wavelength comprises a wavelength to be used by this ONU in case of a failure (link failure, OLT failure, etc.) . There may be one or more backup wavelengths assigned to at least one ONU. The backup-wavelengths may be organized as a sorted list providing a sequence and choice of backup-wavelengths to be selected for allocation purposes. The at least one backup-wavelength could be stored with the ONU or it may be stored centrally for at least one ONU. The at least one backup-wavelength could be provided (set up, amended, deleted, etc.) to the at least ONU via the optical network. The at least one backup-wavelength for at least one ONU may be planned (and/or provided) by a OLT currently con^¬ nected, by a backup OLT or by a central OLT or any other net- work component.

According to a next embodiment, the at least two OLTs provide protection switching and/or load sharing by reserving a certain amount of wavelengths in order to provide in particular - 1+1 redundancy; or

- 1:N redundancy.

The problem stated above is also solved by a device compris^¬ ing a and/or being associated with a processor unit and/or a hard-wired circuit and/or a logic device that is arranged such that the method as described herein is executable thereon .

According to an embodiment, the device is a communication de- vice, in particular a or being associated with an OLT and/or an ONU or any other component of an optical network.

The problem stated supra is further solved by a communication system comprising the device as described herein.

Embodiments of the invention are shown and illustrated in the following figure:

Fig.l shows a symbolic network arrangement in an optical

UDWDM based on a WDM split to 8 optical bands.

In an UDWDM system, each of the m WDM filter's output channels comprises a number n of wavelengths which are split by a passive splitter for the each of n users. For an efficient utilization of the optical fibers the total optical band of a WDM output channel is occupied. This works sufficiently well as long as each of the n users (or subscribers) is provided with the same data rate. If a user needs a higher data rate, two approaches may be ap^¬ plicable : (a) If possible, a modulation format and thus a data rate may be enhanced. However, such a solution requires an increased signal to noise ratio, which sometimes is not available . (b) An additional wavelength band may be coupled to the

splitter thereby increasing the data rate.

Approach (b) also has the advantage that the previous wave^¬ length band is protected, e.g., if one of the coupled wave- length bands fails, there still is a connection (providing a decreased data rate, but still at least some data rate) to and from the user.

Hence, advantageously, 2:n splitters can be used to connect n users via 2 connections directly or indirectly, e.g., via further (CWDM) splitters with OLTs.

Each 2:n splitter may improve the data rate towards its users and it may provide a protection capability as it can be con- nected to two different OLTs via separate lines.

Fig.l shows a symbolic network arrangement in an optical UDWDM network comprising two OLTs 101, 102, each of which being connected to ONUs 108 to 119 via CWDM splitters 103, 104 and passive 2:n splitters 105 to 107. As an example, Fig.l indicates a series of eight 2:n splitters, wherein only three of them are depicted in Fig.l.

The OLT 1 is connected to the CWDM splitter 103 and the CWDM splitter 103 is connected via separate lines to each passive 2:n splitter 105 to 107. The OLT 2 is connected to the CWDM splitter 104 and the CWDM splitter 104 is connected via separate lines to each passive 2:n splitter 105 to 107. The passive 2:n splitter 105 supplies ONUs 108 to 111, the passive 2:n splitter 106 supplies ONUs 112 to 115 and the passive 2:n splitter 107 supplies ONUs 116 to 119. The ONU may also be referred to as user terminal.

Hence, according to Fig.l eight outgoing fibers of each CWDM splitter 103, 104 are connected with eight 2:n passive split^¬ ters 105 to 107, wherein each passive splitter may also be referred to as a power splitter. For example, the first out^¬ put of the CWDM splitter 103 can be connected to the 2:n pas^¬ sive splitter 105, the second output of the CWDM splitter 103 can be connected to the 2:n passive splitter 106 and the eighth output of the CWDM splitter 103 can be connected to the 2:n passive splitter 107.

The mapping of the outputs of the CWDM splitter 104 to the 2:n passive splitters 105 to 107 can be different from the mapping of the outputs of the CWDM splitter 103. For example, the first output of the CWDM splitter 104 is connected to the eighth 2:n passive splitter 107, the second output of the CWDM splitter 104 is connected to the first 2:n passive splitter 105 and the third output of the CWDM splitter 104 is connected to the second 2:n passive splitter 106. In other words, the mapping of the outputs of the CWDM splitter 104 is cyclic shifted by 1 compared to the CWDM splitter 103.

It is noted that such cyclic shift of output mapping is only on examples as how the outputs of the CWDM splitters 103, 104 may be connected with the 2:n splitters 105 to 107. Other kinds of mixing the order of how to associated outputs may be applicable as well.

The CWDM splitter 104 may serve, e.g., as a backup resource for enhancing the data rate and/or for load sharing purposes. Hence, the users or ONUs 108 to 119 may be provided with ad^¬ ditional data rate to cope with bottleneck situations as well as they are protected by an additional resource. Hence, in case of failure of, e.g., OLT 101, the backup OLT 102 may take over at least a portion of the traffic previously con^¬ veyed by OLT 101. Advantageously, the approach provided allows to protect the segment between each 2:n passive splitter and the OLTs. Hence, wherever a failure or bottleneck situation occurs in this segment, the backup resource (OLT 102 in the example of Fig.l) may take over and protect the failed line(s) or fiber (s). Also, additional capacity may be provided by said additional OLT 102.

This approach suggests using (at least) two OLTs in order to provide an efficient protection scheme. However, both OLTs may be connected to a splitter site via disjunctive fibers.

In classical protection schemes, one of the OLTs would be on hot standby, but not transmitting data. In case a signal loss event is detected, the standby OLT is activated and its transmission lasers are switched on. This usually takes some time resulting in a significant delay.

In the proposed schemes, both UDWDM OLTs may continuously transmit light. Each ONU is free to select an OLT thereby al- lowing and/or enabling load sharing between said OLTs. If one OLT fails, the ONUs losing their connection just select a wavelength from the other OLT and get connected to or sup^¬ plied by the other (still active) OLT. As an option, the spectra from the two OLTs may be inter^¬ leaved such that the wavelengths provided are alternating be^¬ tween the OLTs, in particular throughout the whole spectrum.

The approach further comprises the following advantages: a) In case a failure at one OLT or on one fiber is de^¬ tected, the ONU may scan a subsequent adjacent wave^¬ length and allocate such wavelength (if it can be alio- cated) . Thus, switching to an alternative wavelength can be achieved with only a small delay. In particular, re^¬ quirements set forth by some operators to not lose a connection for more than 50ms can be met. b) A CWDM splitter can be deployed within the splitter tree and still the OLTs are be able to serve all CWDM bands.

The protection switch decision can be handled by the ONU and/or by the OLT . If the OLT fails, the ONU may decide to switch to an alternative OLT (or to the other OLT if there are two OLTs) .

If the OLT does not fail, but has lost connectivity to the core network, the OLT may notify ONUs currently connected to this OLT advising the ONUs to switch to the other OLT.

As an alternative, the OLT can trigger protection switching to the alternative OLT in case of maintenance, e.g., in case the OLT hardware is to be repaired or replaced.

Whatever device decides to initiate protection switching, such activity may be visible to the new OLT, because the ONUs are going to register with the new OLT. Therefore the new OLT can take over responsibility for these ONUs that decided (or have been notified) to switch to the new OLT. Hence, the new OLT may inform the core network (e.g., by signaling means) that at least one ONU has moved from the old OLT to the new OLT.

Such signaling may be done via routing protocols or via MPLS label distribution or switching protocols. Hence, protection switching in the OLT-ONU access line can be coordinated with a core-OLT.

If the OLTs operate in an interleaving mode (adjacent wave^¬ lengths being provided by alternating OLTs) , each of two spectrally interleaved OLTs may track the frequencies of the other OLT to avoid collisions due to drifting laser sources.

At least one of the following embodiments may be realized: a) A dedicated or a currently idle first OLT is locked to a wavelength from a second OLT and periodically scans for wavelengths of the second OLT. If the first OLT detects that the second OLT has drifted, the first OLT may lock to the drifted wavelength. Alternatively, the first OLT may be informed to follow the drift of the second OLT. b) A portion of the light from each OLT can be coupled back to the other OLT and in each OLT an independent ONU-like unit may perform a functionality according to the ONU as described under a) above. c) In case of signal loss from the first OLT, each ONU con^¬ nected to that first OLT may (immediately) tune to an adjacent wavelength, which is provided by the second OLT, thus enabling a significantly reduced switching de^¬ lay .

Also, a protection speed switching from one OLT to the other OLT can be optimized. During normal operation, both OLTs are active. In such state, e.g., as a matter of precaution, for each ONU (or a selection thereof) a "in case of"-wavelength of the redundant OLT can be assigned ("pre-planned assign^¬ ment") . If the first OLT fails, all ONUs know exactly, which wavelengths to be allocated. This may systematically and/or statistically reduce any conflicts that may otherwise stem from the fact that all (or several) ONUs at the same time have to search for (identical) free wavelengths thereby col^¬ liding with one another.

This pre-planned assignment of redundant wavelengths can be done in the active OLT, in the redundant OLT or in a central^¬ ized management/planning system. In all cases the redundant wavelengths may be configured to the ONU via an actual active connection between the active OLT and the ONU.

As an example of implementation, this protection scheme may utilize at least two OLTs. Preferably, half of the each OLT ' s channels are not used in order to achieve a "1+1 redundancy". It is also possible to use more than two OLTs thereby ena^¬ bling a "1:N redundancy". For example, three OLTs may use 67% of the OLT channels during normal operation, thereby reducing the cost.

List of Abbreviations:

CWDM Coarse WDM

OLT Optical Line Terminal

ONU Optical Network Unit

PD Photo Diode

PON Passive Optical Network

UDWDM Ultra Dense WDM

WDM Wavelength Division Multiplex

Claims

A method for data processing in an UDWD network,

- wherein one of at least two OLTs communicates with at least one ONU,

- wherein the at least one ONU is connected to a

splitter that is connected to each of said at least two OLTs,

- wherein the at least two OLTs are active, and

- wherein the at least two OLTs provide protection

switching and/or load sharing.

The method according to claim 1, wherein the splitter is connected via disjunctive fibers to each of said at least two OLTs.

The method according to any of the preceding claims, wherein several ONUs are connected to the splitter or to an additional splitter, wherein the additional splitter is connected to each of said at least two OLTs.

The method according to any of the preceding claims, wherein the splitter is a m:n splitter, in particular a passive m:n splitter, wherein m refers to a number of OLTs that can be connected and n refers to a number of ONUs that can be connected.

The method according to any of the preceding claims, wherein the splitter is connected to the at least two OLTs via at least two C DM splitters, one CWDM splitter per each OLT .

The method according to any of the preceding claims, wherein each of the at least one ONU selects one of the at least two OLTs. The method according to any of the preceding claims, wherein wavelengths supplied by the at least two OLTs are at least partially interleaved.

8. The method according to any of the preceding claims, wherein said protection switching and/or load sharing is triggered by at least one OLT and/or by the at least one ONU.

9. The method according to any of the preceding claims, wherein at least one OLT tracks wavelengths of at least one other OLT, in particular to avoid collisions due to drifting laser sources.

10. The method according to claim 9, wherein the at least one OLT follows a wavelength drift of at least one other OLT.

11. The method according to claim 9, wherein the at least one OLT tracks a wavelength drift of the at least one other OLT and informs this at least one other OLT to adjust its wavelength.

12. The method according to any of claims 10 or 11,

- wherein the at least one OLT is optically coupled to the at least one other OLT,

- wherein the at least one other OLT performs an

assessment of a wavelength drift based on the optically coupling and adjusts its wavelength based on the assessment.

13. The method according to any of the preceding claims, wherein the at least one ONU is associated with at least one backup-wavelength.

14. The method according to any of the preceding claims, wherein the at least two OLTs provide protection switching and/or load sharing by reserving a certain amount of wavelengths in order to provide in particular

- 1+1 redundancy; or

- 1:N redundancy.

15. A device comprising a and/or being associated with a processor unit and/or a hard-wired circuit and/or a logic device that is arranged such that the method according to any of the preceding claims is executable thereon .

16. The device according to claim 14, wherein said device is a communication device, in particular a or being associated with an OLT and/or an ONU.

17. Communication system comprising the device according to claim 15.