WO2008117032A1

WO2008117032A1 - Processing data

Info

Publication number: WO2008117032A1
Application number: PCT/GB2008/001003
Authority: WO
Inventors: Peter Healey; Vincent Robert Chabasseur
Original assignee: British Telecommunications Public Limited Company
Priority date: 2007-03-26
Filing date: 2008-03-20
Publication date: 2008-10-02

Abstract

The present invention is concerned with processing data, in particular data from a plurality of terminals time division multiplexed onto a common optical path. There is provided a method of operating a repeater in an optical network having a head end that transmits a signal stream, which signal stream includes instructions to a plurality of terminals for controlling the timing of traffic from the terminals such that data bursts from the different terminals are time division multiplexed onto a common optical path, the method including the steps of: (i) monitoring the signal stream from the head end and reading the timing instructions; (ii) from the instructions, inferring the identity of the terminal responsible for a given arriving data burst along the common optical path; and, (iii) processing the given data burst in dependence on the inferred identity. Because the repeater is arranged to monitor the signal stream from the head end and read the timing instructions therein, account can be taken of the particular type of distortion that is to be expected in the arriving data, allowing the data to be more effectively compensated that it would be were the data processed independently of its distortion.

Description

Processing data

The present invention is concerned with processing data, in particular data from a plurality of terminals time division multiplexed onto a common optical path.

In a so-called Passive Optical Network PON, data from a plurality of terminals is passively time division multiplexed at a branch junction before reaching a head end. In order to extend the reach of the network, it is helpful to introduce at least one repeater in the network. The repeater usually compensates for signal attenuation and distortion from the head end, allowing these signals to travel further. Usually, the repeater is positioned at a point between the junction and the head end, so that data to or from a plurality of terminals can benefit from the repeater (although if more than one branch junction is provided, a repeater may be positioned between two branch junctions).

A repeater may be arranged to compensate for distortion of signals travelling from the terminals, in addition or alternatively to compensating signals from the head end. However, the terminals may be situated at different optical path distances from the repeater, with the result that the data from the different terminals will have suffered different distortion.

According to one aspect of the present invention, there is provided a method of operating a repeater in an optical network having a head end that transmits a signal stream, which signal stream includes instructions to a plurality of terminals for controlling the timing of traffic from the terminals such that data bursts from the different terminals are time division multiplexed onto a common optical path, the method including the steps of: (i) monitoring the signal stream from the head end and reading the timing instructions; (ii) from the instructions, inferring the identity of the terminal responsible for a given arriving data burst along the common optical path; and, (iii) processing the given data burst in dependence on the inferred identity.

Because the repeater is arranged to monitor the signal stream from the head end and read the timing instructions therein, account can be taken of the particular type of distortion that is to be expected in the arriving data, allowing the data to be more effectively compensated that it would be were the data processed independently of its level of distortion. Further aspects of the invention are specified in the claims. The invention will now be further described, by way of example only, and with reference to the following drawings in which:

Figure 1 shows an optical network in accordance with the present invention; Figure 2 shows a downstream frame structure for communication in the optical network of figure 1 ;

Figure 3 shows an upstream frame structure; Figure 4 is a block diagram of a repeater for use in the optical network of figure 1 ; Figure 5 shows a further optical network;

Figure 6 shows in more detail functional blocks in the repeater of figure 4; and, Figure 7 is a block diagram of an optical compensation module.

Figure 1 shows a passive optical network (PON) 10 in which a head end (or OLT) 12 having an optical source 13 broadcasts, in a downstream direction, information in the optical domain to a plurality of optical network units (also known as ONU, outstation or terminal devices) 14 over an optical distribution network (or ODN) 16. The head end 12 also receives information from the network units 14 in the upstream direction. The ODN is formed in a tree like structure, having a trunk optical fibre 18 connected to a passive optical splitter 20, which splitter distributes the incoming light from the trunk fibre 18 to a plurality of branch optical fibres 22. In the example of figure 1 , the splitter 20 has an 8- way split, and couples light from the trunk fibre to eight branch fibres, although only two are shown for clarity.

Each branch fibre from the splitter 20 is connected to a respective further splitter 24, which further splitter distributes the light from each branch to a plurality of further branches 26, each further branch being connected to a respective network unit. Each further splitter has a split of 32, such that the head end 12 is connected to 256 network units 14. Yet further levels of split (not shown) may be provided so that a greater number of network units can be connected to the head end.

The head end has a scheduler stage 38 for controlling the timing of signals from the respective network units 14, such that there is essentially no risk (or a reduced risk) that upstream signals from one network unit will collide with signals from another network unit where optical paths are combined at a splitter. The network units are configured to transmit short bursts of data in response to scheduling instructions from the head end, the scheduler stage 38 of the head end being arranged to perform a scheduling function such that the transmission of bursts are timed so as not to overlap. Hence, data from the different OLTs is passively interleaved or equivalently multiplexed (in this example temporally, that is, in a time-division manner), in the normal way of a PON. The head end can then access data from each network unit using a TDMA (Time Division Multiplexed Access) protocol.

The scheduling instructions from the head end contain an instruction for a given network unit to transmit data for a predetermined time interval at a predetermined time, with respect to a centralised time (although other in protocols data transmission is timed with respect to the local arrival time of a control signal). Accordingly, each network unit has a respective timing unit 40 for measuring the time at which a data burst is to be transmitted in response to a control signal from the head end. Each network unit 14 will normally be connected to customer equipment such as telephony equipment or other communications equipment 42, and will be arranged to buffer data from the customer equipment in order to transmit it at the correct time in the upstream direction.

The network units each have an address or identity associated therewith issued by the head end, and the head end is arranged to transmit broadcast instructions which associate given data with a given address. Each network unit is arranged to monitor the broadcast information from the head end, to capture data addressed to it, and to forward the captured data towards the correct customer equipment.

Data in the upstream and downstream directions is transported as a series of cells arranged in a frame structure. An example of a frame structure for downstream traffic transmitted by the head end is shown in figure 2. A frame 50 includes a plurality of sequentially arranged data cells or time slot 52, each of which has a data payload 54 for carrying traffic data intended for costumer equipment, and a header 56, which header includes protocol-specific fields for defined types of information, such as an identifier field for the identity of the network unit for which a given data cell is intended (each network unit is arranged to forward data in the payload to customer equipment, whereas data in the head is not normally so forwarded). Signalling cells 58 are provided between some of the data cells at predetermined positions within a frame (in the present example, two signalling cells and 54 data cells are provided per frame). The signalling cells (also known in some protocols as PLOAM or "Physical Layer Operation, Administration and Maintenance" cells) contain the scheduling instructions (also know as "grants") from the head end which permit one or more specified network units to transmit a specified number of upstream data cells at a respective specified time. The signalling cells also contain dedicated fields for synchronisation information to allow the network units to achieve synchronisation. Additional fields are provided for ranging, error control, security and information relating to maintenance functions. Examples of known protocols for governing signalling, traffic transmission and other aspects of network operation include the ITU standards G983.1 and G983.4.

In the upstream direction, the network units are instructed, by means of downstream signalling cells, to each transmit upstream data cells at an appropriate time so that the cells from the different network units come together at a junction in an interleaved fashion so as to form a downstream frame structure as shown in figure 3, guard bands 59 being provided between the cells to allow for timing irregularities. Consecutive cells may or may not originate from the same network unit. Although the frame structure of Figure 3 is shown as a series of consecutive cells, some cells may be omitted, for example if one or more network units is not instructed to transmit data. The signal stream formed by consecutive frames may appear as spaced apart data burst (each formed by one or more cells), rather than a continuous signal.

In a similar fashion to the downstream data cells, the upstream data cells also each have a payload 54 and a header 56. However, the protocol-specific fields in the upstream direction will normally be different to those in the downstream direction, and include respective fields for; a network unit serial number; inscription keys; and, parity checking (error correction). In addition, the header includes one or more management fields for management information.

In order to compensate for losses and other degradation of data, and thereby extend the reach of the PON, a repeater 100 is provided along the path of the trunk fibre 18; that is, between the head end 12 and the first downstream junction 21. The repeater is located a significant distance from the head end, for example 50km or even 100km therefrom. A more detailed view of the repeater is shown in figure 4. The repeater comprises a repeater module 102, which operates largely as a conventional repeater, and a management module 104 for performing management functions with regard to the repeater module 102. The repeater module 102 includes an electronic module 106, and respective upstream and downstream opto-electrical interfaces 108, 110, each for converting incoming optical signals into the electrical domain and for converting electronic signals from the electronic module 106 into optical signals (that is, the upstream interface receives and transmits optical signals from and to the head end, whilst the downstream interface receives and transmits signals from and to the network units). In operation, downstream signals from the head end are converted into the electrical domain and processed by the electronic modulator 106 before being converted back into the optical domain and transmitted towards the network units. The electronic module is arranged to; re-shape; re-time; and, re-amplify signal pulses (or at least perform some of these tasks) in order to compensate for loss and distortion of the signal as it travels along the optical fibre path. In this way, the repeater module reproduces, repeats or otherwise regenerates at least in part the information content of optical signals. The electronic module 106 also performs a regeneration function in respect of downstream traffic, but as part of this regeneration, is arranged to at least in part even out the amplitude of data from the different network units, in order to compensate for the fact that the signal from different network units is likely to have suffered different amounts of attenuation, due to their different path distance to the repeater.

The management module 104, which is electrically connected to the repeater module 102, is arranged to monitor certain attributes relating to the management or operation of the repeater module, and to report the results as management data to the head end. Examples of management data include details of one or more of the following: incoming optical power from the network units; incoming optical power from the head end; power level of signals transmitted to the network units; power level of signals transmitted to the head end; power supply voltage(s); temperature at one or more locations; and, error performance. Additionally or alternatively, the management data may include an alarm signal generated in dependence on the values of one or more of the monitored attributes.

The management module 104 is arranged to communicate with the head end using the same communication protocol as that used by the network units. In particular, with regard to communication in the downstream direction, the management module is arranged to monitor a copy of the signal stream in the electrical domain and read data cells addressed to it, in an analogous fashion to the way in which the network units each receive a copy of the signal stream in the optical domain. For communication in the upstream direction, the management module 104 is arranged to insert data cells into the data stream in the electrical domain, correctly timed in accordance with scheduling instructions from the head end, again in an analogous fashion to the network units. Thus, the management module 104 can be viewed as a virtual network unit or ONU, allowing existing functionality at the head end to be employed in order to communicate with the repeater.

The management module 104 includes: a memory 112 for storing management data obtained as a result of the monitoring of the repeater module; an input 114 for receiving a copy of the downstream signal stream from the head end; an output 116 for inserting data into the upstream signal stream; a timing stage 118 for controlling the timing of data onto the upstream signal stream; and, a controller stage 120 for controlling the operation of the management module 104. The controller stage 120 is operatively coupled to the electronic module 106 and/or one or both of the upstream and downstream opto-electrical interfaces 108, 110 in order to perform the required monitoring functions necessary to collect the management data for storage in memory 112.

The controller stage 120 is configured to read the signalling cells from the head end, and in response to scheduling instructions in a signalling cell, to transmit one or more data cells at a time specified by the scheduling instructions. The timing of the transmission of the data cells is governed by the timing stage 118, which includes a clock unit 119 synchronised to a master clock at the head end using the transitions in the data stream and synchronisation information in the signalling cells. The frame synchronisation is achieved by the controller stage through the monitoring of frame synchronisation signals in the signalling cells.

The management module has a short term buffer 121 for delaying the transmission of data by an adjustable offset delay time. The controller stage 120 is responsive to offset signals in the signalling cells so as to adjust the buffer delay to a delay time specified in the offset signals. The real network units 14 also each have an adjustable buffer (not shown). This allows the head end to adjust the adjustable buffer of each network unit and likewise that of the management module so as to take into account the different path lengths and hence transit times from each network unit and the repeater. In a set-up phase, a ranging function is normally carried out to calculate the required value of offset delay time for each network unit as well as for the repeater.

In the ranging function carried out in respect of the repeater, a ranging instruction is broadcast with an address corresponding to that of the repeater, in response to which the management module sends a return message. The head end then calculates the round trip time from the arrival time of the return message, and hence determines the required offset value for the management module of the repeater (from the round trip time, the optical path distance of the repeater from the head end can also be inferred). A corresponding ranging function is performed in respect of each network unit.

The control stage is also responsive to buffer adjustment signals (transmitted within a predetermined field of the signalling cells) which instruct the controller stage to adjust the buffer delay so as to compensate for small variations in transit time during normal operation, and thereby maintain synchronisation.

In order to transmit management data, the control stage 120 is arranged to extract the required data from the memory 112 and pass this data to the short term buffer 121 , from which it is inserted into the header of one or more data cells, in particular into the protocol- specific management field(s) of the header. Although the management data could simply be inserted into the payload of a data cell, by inserting this data into management-specific fields, use can be made of existing management functionality associated with the head end for managing the real network units.

In addition to the protocol-specific instructions from the head end, the management module may also receive supervisory or other maintenance instructions in one or more data cells addressed to it (the maintenance instructions need not be specific to the operating protocol of the PON). The maintenance instructions may for example instruct the repeater to alter the optical power output to the network units, and/or to the head end, or control other attributes of the repeater, operation. The repeater may include a redundant electronic module or redundant parts thereof, or other components such as a redundant opto-electronic interfaces for fault tolerance. In such a situation, the management module will be configured to perform a switching function, switching the repeater in response to fault instructions from the head end such that the repeater uses one or more of the redundant components. The management module is configured to respond to other instructions from the head end in accordance with the operating protocol of the PON. For example, the management module has a serial number permanently stored therein (for example on a read-only memory written during manufacture), which, in an initialisation phase, the management module will transmit to the head end, in order that the head end can allocate to the management module an identifier (which identifier the management module commits to memory) so that the management module can be addressed by the head end using the identifier, for example when being granted upstream bandwidth through scheduling instructions from the head end, in a similar fashion to the network units.

During normal operation of the PON, the management module will appear to the head end like each of the other network units, and thus can be invited to transmit data using the scheduling instructions in the signalling cells. Thus, when management data from the management module is required, the management module can be schedule in the normal way to transmit one or more data cells in otherwise empty time slots in the frame structure of figure 3.

Although in the above embodiment a synchronous PON system is described, protocols relating to asynchronous systems are also possible. For example, the head end may simply send instructions for a network unit to transmit data after a specified delay, which delay is measured from the time at which the network unit receives the instructions, rather than with respect to an absolute time. Of course, the management module will need to be adapted to the particular protocol in use with a given PON.

The repeater 100 may be used as a wavelength shifter, for example by arranging the opto-electronic interfaces 108, 110 such that light received at one wavelength by one of the upstream and downstream interfaces is transmitted at a different wavelength by the other of the upstream and downstream interface.

In figure 5, there is shown an optical network comprising a first, second and third PON

10a, 10b, 10c, which share a common trunk fibre 18 through a Wavelength Division

Multiplexing (WDM) arrangement, in which the trunk fibre carries light for each of the PONs at a respective first, second and third wavelength. The head ends 12a, 12b, 12c of the respective PONs are arranged to transmit at the same wavelength, and, likewise, the network units 14 of the different PONs are also arranged to received light at a common wavelength. Therefore, a respective wavelength shifter 200a, 200b, 200c (configured essentially as that of figure 4) is provided for converting signals at the common wavelength from each of the head ends into corresponding signals at a respective one of the first, second and third wavelength. Light from the different wavelength shifters 200a, 200b, 200c is combined at an upstream WDM junction 202 and transmitted over the trunk fibre 18 before being demultiplexed at a downstream WDM junction 204 into respective signals at the first, second and third wavelengths. Signals at the first, second and third wavelengths are passed to the downstream wavelength shifters 100a, 100b, 100c over respective local fibres 206a, 206b, 206c, where the signals are converted to a common wavelength (normally the same common wavelength as output by the head ends) and fed to at least one local splitter 20a, 20b, 20c for distribution to respective groups 14a, 14b, 14c of network units belonging to the first, second and third PON respectively.

In this embodiment, the use of the repeater of figure 4 as a wavelength shifter allows existing head end and network unit equipment to be used when wavelength multiplexing part of the optical path of a plurality of PONs over a common wave guide. Clearly, depending on the relative positions of the upstream and downstream wavelength shifters, the wavelength shifters may or may not each be needed to perform the re-shaping, re- amplification and re-timing normally performed by a repeater.

Figure 6 shows in further detail one embodiment of the repeater 100 of figure 4, in which components corresponding to those of figure 4 are given corresponding numerals. The repeater module 102 is based on an existing repeater module (such as the ZB2 G5REPT made by Zenko) having a clock and data recovery unit 209 for recovering the clock signal and data in the downstream signal stream, and a processing unit 211 for re-generating upstream data using the clock signal from the clock and data recovery unit 209. However, instead of downstream data from the interface 110 simply being fed to the processing unit as is the case with a conventional repeater, an OR function is performed at an OR gate 213 on the downstream data with data from the management module 104. This allows data from the management module to be inserted into the signal stream in the upstream direction. The management module 104 is connected to the clock and data recovery unit 209 in order to obtain a copy of the downstream continuous wave signal. A delay stage may be provided between the management module and the processing unit to simulate the delay expected with a real terminal due to the additional optical fibre length (and the consequent transit time) downstream of the repeater.

Clearly, one advantage of the above embodiment is that the management module is easily appended to an existing repeater module, with only minor modifications being required. However, the one or more components of repeater module and of the management module may be integrally formed, for example on a common card or chip device. In particular, some of the electronic functionality of the repeater module and the management module may be implemented on a common processor chip.

In a further embodiment, the electronic module 106 includes a compensation module 144 shown schematically in figure 7. The compensation module 144 has processor facility 150 having at least one processor on which runs a compensation algorithm for treating or otherwise equalising data so as to correct the data received from the network units for any inter symbol interference (ISI) or other distortion. The compensation algorithm has a plurality of adjustable characteristics each of which is representative of an aspect of the way in which the compensation algorithm treats data. The adjustable characteristics are governed by a set of coefficients (parameters) such that each co-efficient is associated with a respective characteristic. The choice of values for the coefficients will depend on the degree of distortion, and on the nature (type) of the distortion. The nature of the distortion may depend on several factors, such as the distance signals have travelled along an optical fibre, the material characteristics of the fibre, the bit rate and possibly the ambient conditions local to the fibre.

The distance between the central station and an network units will normally differ from network unit to network unit, the difference being at least 20km, 50km or even potentially

100km or more. Thus, the nature and extent of the distortion is likely to be different for data from different network units. Furthermore, although the distortion in incoming data at the central station is likely to change from cell to cell (i.e., as fast as about every 424 bits if there are 424 bits per cell), the distortion is likely to change only slowly, if at all, in respect of data from a given network unit.

, In order to allow the compensation module to be able to predict which network unit data will arrive from, the controller stage 120 of the management module 104 is arranged to capture the scheduling instructions from the signalling cells in respect to each of the network units (rather than just those in respect of the repeater). To do this, the compensation module will have knowledge of the particular protocol used by the head end in order to identify the signalling cells and the scheduling instructions therein. In particular, the compensation module has knowledge of the format of the scheduling instructions, including the identifier for identifying each outstation which is to transmit data, and timing information indicating when an outstation is to transmit that data. From the captured identifier and timing information, the compensation module can infer the identifier associated with data returned in response to scheduling instructions, and hence infer the type of distortion to which the data has been subjected. The data can then be categorised in dependence on the inferred type of distortion. Although an identifier in the scheduling instructions is read in order to allow the incoming data to be categorised, the identifier will normally have been allocated by the head end, and the real identity of a network unit (as defined by its permanent serial number for example) need not be known.

The compensation module 144 is connected to the controller stage so as to receive the scheduling instructions, and stores these in a local memory 152, such as a fast access memory or RAM. The compensation module then uses the stored scheduling information in order to make the prediction as to which network unit data is going to arrive from.

In addition to the captured scheduling information, the compensation module stores a respective set of coefficients in respect of each network unit, in the form of a table, containing in one row a set of identifiers one for each network unit, and in another row, the respective sets of coefficients, the table containing mapping information which maps each network unit identifier to a set of coefficients. From the stored scheduling information, the compensation module infers the identity of the network unit from which each arriving cell of data originates. Each time it is inferred that data will arrive from a different network unit, the compensation module retrieves the set of coefficients stored for that network unit, and runs the algorithm in accordance with the newly retrieved coefficients so as to treat data from the current network unit. Thus, as a result of the scheduling process, and in particular the scheduling information obtained from the downstream signal by the management module 104, the compensation module 144 knows in advance which network unit the next cell to be received will originate from. The compensation module can therefore quickly switch between the appropriate coefficients to compensate the distortion of the next arriving cell. As a result, adjacent cells can have very different amounts of distortion, as the compensation algorithm may not need to start an optimisation process from the beginning for each cell, but can quickly recall the appropriate set of coefficients and therefore maintain appropriate values in the compensation algorithm.

Clearly, in this embodiment, the management module 104 and in particular the control stage thereof is not simply configured to perform functions corresponding to those of a network unit, but is instead arranged to read the scheduling instructions for each of the network units.

One example of a compensation algorithm is known as a Transversal Filter process. In general terms, the operation of the process involves taking samples at a plurality of points, that is, time positions or "taps" along an incoming stream of data. The taps may be at intervals of less than one bit spacing in the incoming data stream, for example every half or one quarter of a bit period. Furthermore, the sampling intervals need not be regular, and could be irregular, for example to take into account the complexity of arriving data. Preferably, the taps are at successive or "neighbouring" bit positions. For each target bit to be compensated, weighted samples of bits in the neighbourhood of the target bit can then be mixed with the target bit. In particular, the weighed neighbourhood bits are normally added or subtracted from the target bit. (Because of the distortion or overlap between bits, the bits no longer have a 0 or 1 value, but can have values in a continuous range).

The neighbourhood bits will preferably include at least one immediate neighbourhood of the target bit, such that for a 3 tap filter process, the target bit and its two (in this example trailing) neighbours in the data stream are sampled. For a 5 bit tap, the next 4 nearest trailing neighbours of the target bit will be included in the neighbourhood being sampled, etc... To weight the sampled bits, a weighting function is applied to each respective bit, the weighting function for each bit being governed by a respective one of the coefficients in the set corresponding to the network unit whose data is to be corrected. Preferably, the weighting function is a simple factoring function (e.g., multiplication).

To operate the filter process, the compensation module will store a given number of successively arriving bits in respective memory locations in the memory 152, for example in a shift register within the memory 152 (the compensation module works on the analogue signal from the receiver and the stored data is a digital representations of the analogue signal). Essentially, in operation, (i) each successive data bit is weighted by the coefficient associated with the memory location in which that data bit is present, (ii) the weighted values are saved, and, (iii) the data is then shifted, such that data in one memory location is replaced by the data in the immediately trailing bit slot, that is the slot next to have arrived. Step (i), (ii) and (iii) are repeated in order, with the result that for each cycle a value is generated which is a combination of the sample data at each memory location. In particular, for each target data bit, the combination comprises that target data bit together with the weighted bits which trail the target bit in a specified neighbourhood.

Claims

1. A method of operating a repeater in an optical network having a head end that transmits a signal stream, which signal stream includes instructions to a plurality of terminals for controlling the timing of traffic from the terminals such that data bursts from the different terminals are time division multiplexed onto a common optical path, the method including the steps of: (i) monitoring the signal stream from the head end and reading the timing instructions; (ii) from the instructions, inferring the identity of the terminal responsible for a given arriving data burst along the common optical path; and, (iii) processing the given data burst in dependence on the inferred identity.

2. A method as claimed in claim 1 , wherein the step of processing a data burst includes: (i) running a compensation procedure controlled by a set of parameters, each parameter being able to assume a respective value; (ii) storing a set of parameter values in respect of each terminal; and, (iii) selecting a set of parameters for use in controlling the compensation procedure, the set of parameter values being selected in dependence on the inferred identity.

3. A method as claimed in claim 2, wherein a given data bursts has a plurality of successive time positions associated therewith, and the compensation procedure is arranged to compensate data at a given target time position by: (i) sampling the value of the data at the target time position; (ii) sampling the value of the data at time positions in the neighbourhood of the target time position; (iii) weighting, with a respective weighting function, at least some of the data values sampled in step (ii); and, (iv) mixing the weighted data values sampled in said neighbourhood with the data value sampled at the target time position.

4. A method as claimed in claim 3, wherein the weighting function for each time position is governed by a respective one of the parameters in the parameter set corresponding to the network unit whose data is to be compensated.

5. A method as claimed in any one of the preceding claims, wherein the signal stream from the head end is arranged in a frame structure in which a given frame has a plurality of data cells and one or more signalling cells, the timing instructions residing in the signalling cell(s).

6. A method as claimed in any one of the preceding claims, wherein the signal stream is broadcast by the head end.

7. A method as claimed in any one of the preceding claims, wherein the optical network is a passive optical network.

8. A repeater suitable for use in repeating optical signals in an optical network having a head end that transmits, in a downstream direction, a signal stream including commands to a plurality of terminals for controlling the timing of traffic from the terminals in an upstream direction such that data bursts from the different terminals are time division multiplexed onto a common optical path, the repeater having: means for receiving the downstream signal stream; means for monitoring, in use, the downstream signal stream so as to read the timing commends therein; means for receiving, in use, the upstream time division multiplexed data from the different terminals; means for inferring the identity of the terminal responsible for an arriving data burst on the basis of the timing commands; and, means for processing the data from a terminal in dependence on the inferred identity of the responsible terminal.

9. A repeater as claimed in claim 8, wherein the means for processing the data are arranged to carry out a compensation procedure for compensating for distortion of the data.

10. A method of operating a repeater in an optical network having a head end that transmits a signal stream, which signal stream includes instructions to a plurality of terminals for controlling the timing of traffic from the terminals such that data bursts from the different terminals are time division multiplexed onto a common optical path, the method including the steps of: (i) monitoring the signal stream from the head end and reading the timing commands; (ii) categorising data along the common optical path independence on the identity of the terminal from which the data originates; (iii) from the timing commands, inferring the category of the terminal responsible for an arriving data burst; and, (iv) processing the data from a terminal in dependence on the inferred category of the data.