HK1078402A - Optical network communication system - Google Patents

Description

Optical network communication system

Technical Field

The present invention relates generally to information transmission in communication networks, and in particular to a protocol for transmission over a passive optical network.

Background

A point-to-multipoint Passive Optical Network (PON) operates as a communication system by broadcasting optical signals downstream from a central unit, referred to herein as an Optical Line Terminal (OLT), to Optical Network Terminals (ONTs). The signal is transmitted from the OLT to the ONTs via a fiber optic cable and a passive optical splitter, which includes the physical structure (fabric) of the network. For upstream communications, each ONT must be able to transmit signals that are not interfered with by the other ONTs. One of the known methods in the art for performing such upstream and downstream transmissions is to utilize Time Division Multiple Access (TDMA), wherein each ONT is assigned a window when only it is capable of transmitting, and wherein the OLT also has windows for transmissions to specific ONTs. Other methods for avoiding interference include transmitting signals at different wavelengths using Wavelength Division Multiple Access (WDMA). Combinations of TDMA and WDMA are also known in the art.

Us patent 5,173,899 to Ballance, the contents of which are incorporated herein by reference, describes a method for communicating in a passive optical network. The OLT sends downstream TDM frames, frames including information, synchronization signals, and grants to downstream ONTs (sending upstream). The ONT transmits an upstream TDM signal in response to the grant and the synchronization signal.

U.S. patent No. 5,355,368 to Dore et al, the contents of which are incorporated herein by reference, describes a method for allocating time slots in a TDMA point to multipoint network. The network operates in a half duplex manner, i.e. the network terminals alternate between transmitting and receiving. This approach reduces the "dead time" required between adjacent downstream time slots of the OLT, which is the round trip time in the network for transmissions between the OLT and the ONTs. This reduction is achieved by: when a first ONT is receiving information addressed to a second ONT, the first ONT is given authorization to transmit.

5,515,379 to Crisler et al, the contents of which are incorporated herein by reference, describes a time slot allocation system in a communication system. The communication unit sends a first data packet requesting permission to transmit to the time slot allocator. The data packet contains a request to allocate a plurality of time slots or a request to transmit a plurality of data packets. In either case, the allocator allocates consecutive time slots to the units for transmitting their data packets.

U.S. patent 5,528,592 to Schibler et al, the contents of which are incorporated herein by reference, describes a method for routing Asynchronous Transfer Mode (ATM) cells. (the data packet is comprised of a plurality of ATM cells.) the method includes receiving cells in the route cell buffer corresponding to the beginning and end of the packet. The router determines the routing information of the packet by these units. The information includes a routing label that determines an output port of the packet and an identifier of a switching channel that determines a source of the packet and a destination of the packet.

U.S. patent No. 5,838,687 to Ramfelt, the contents of which are incorporated herein by reference, discloses a method of reusing timeslots in a dynamic synchronous transfer mode (DTM) segmented network. Access to a slot is controlled by a slot token and writing to a slot is performed only by the controller owning the corresponding token for that slot. A block token (blocktoken) is used to represent a group of tokens in a single control message. The method includes extending the DTM block token format to include parameters describing the segmentation between the source and destination nodes. The block token capacity is reserved only on segments between nodes and enables synchronous transmission in the same time slot over separate segments of the network.

United states patent No. 5,982,780 to Bohm et al, the contents of which are incorporated herein by reference, describes centralized and distributed management of communication resources in a DTM network. In a centralized approach, a server node assigns tokens corresponding to time slots to unidirectional data flows on the communication links. If the server has available capacity, it reserves and transmits tokens to other nodes on the link according to requests from those nodes. In a distributed scheme, the functionality of the server is spread among two or more nodes connected to the link.

Disclosure of Invention

It is an object of some aspects of the present invention to provide a method for transmitting signals in a communication network.

Other objects of some aspects of the present invention are to provide a method for transmitting a TDM signal in a Passive Optical Network (PON).

In a preferred embodiment of the present invention, an Optical Line Terminal (OLT) transmits optical signals downstream to a plurality of Optical Network Terminals (ONTs). The OLT is coupled to the ONTs via a passive optical distribution structure, thus forming a PON. The OLT acts as a controller for downstream signals as well as signals sent upstream by the ONTs and received by the OLT. The downlink signals are transmitted in frames having a constant period, and each downlink frame includes a plurality of "slots". Each time slot is a fixed number of bytes and within each frame the OLT allocates time slots directed to each ONT in a dynamic manner. The time slot allocation for each frame is implemented in accordance with the amount of data to be sent to or received from the ONTs, as determined by the OLT. Most preferably, the time slot allocation for each downlink frame is performed at substantially the same rate as the frame is transmitted. By allocating a different number of time slots to each downstream frame or upstream window (explained below), the OLT can efficiently configure each frame or window in a variable TDM manner, allocating a variable length of time to each ONT receiving data.

Upstream signals are sent in a TDM fashion from a single ONT in a window that is sent according to a time pointer assigned by the OLT.

Both the upstream and downstream signals include data transmitted in accordance with one or more services, which may operate separately in accordance with disparate protocols. These services typically include constant bit rate services as well as packet-based services. The data transmission of the preferred embodiment of the present invention is independent of the type of service on which the data is sent.

Upstream and downstream signals are transmitted between the OLT and the ONTs via channels that are mapped according to the service according to a one-to-one mapping. The channel and channel parameters, such as bandwidth, are assigned by the operator of the PON at initialization or during operation of the PON. Each channel uses time slots that are allocated according to the bandwidth requirements of the service of that channel.

Data for a particular channel in a downlink frame may be distributed discontinuously within the frame. In addition, data for a particular channel in an uplink window may be distributed discontinuously within the window. Having the special channels in the upstream window and the downstream frame be discontinuously set can significantly enhance the flexibility and efficiency of transmission of these signals compared to systems that do not allow discontinuous transmission.

There is thus provided in accordance with a preferred embodiment of the present invention a method for downstream communication from a central transmission point to a plurality of receiving end points by time division multiplexing of a sequence of frames, each frame being divided into a plurality of time slots, the method including:

receiving at the central transmission point data for transmission to the end points, the data comprising at least a first amount of first data for transmission to a first end point of the plurality of end points and a second amount of second data for transmission to a second end point of the plurality of end points, such that the first and second amounts are variable from each frame to the next frame in the sequence;

allocating a first number of timeslots in each frame to transmit first data to the first endpoint and a second number of timeslots to transmit second data to the second endpoint in response to the first and second amounts, such that the first and second numbers are variable from each frame to the next frame in the sequence in response to changes in the first and second amounts of data; and

the data is transmitted from the central transmission point to the end points during the allocated time slots.

Preferably, the central transmission point comprises an Optical Line Terminal (OLT) and the end points comprise Optical Network Terminals (ONTs), wherein the OLT and the ONTs are operated as transceivers in a passive optical network.

Preferably, the data comprises data sets communicated via respective different industry standard services.

Preferably, the plurality of end points comprises sets of end points operable under different wavelength groups, and the data comprises respective sets of data communicated between the central transmission point and the respective sets of end points via the different wavelength groups.

Preferably, the data comprises a data set communicated via respective different channels, wherein each channel transmits data via a service coupled with the central transmission point and at least one end point.

It is further preferred that the method comprises allocating a respective bandwidth to each channel, wherein allocating the first and second numbers of time slots comprises allocating the first and second numbers of time slots in response to the bandwidth of each channel.

Preferably, allocating a respective bandwidth to each channel comprises changing the respective bandwidth to a different bandwidth in response to a request received by the central transmission point.

It is further preferred that the first and second number of time slots are allocated in response to respective first and second data parameters stored in a memory comprised in the central transmission point.

Preferably, the sum of the first number and the second number of time slots is less than or equal to the bandwidth of each frame sequence.

It is further preferred that the data comprises one or more further quantities of data for transmission to one or more further endpoints of the plurality of endpoints, such that the one or more further quantities may be variable from each frame to the next frame in the sequence, and including allocating a respective one or more further numbers of time slots to be transmitted to the one or more further endpoints respectively.

Preferably, the period of each frame sequence is substantially constant.

Preferably, each frame sequence comprises a header comprising a respective window parameter for each of the plurality of endpoints, each window parameter comprising a time and a size of a window of upstream data each of the plurality of endpoints is allowed to transmit to the central transmission point.

It is also preferred that the respective window parameters are assigned by said central transmission point so that the windows do not collide at the central transmission point.

Preferably, the data comprises at least a third amount of third data for transmission to the first endpoint, such that the third amount is variable from each frame to the next frame in the sequence, and comprises: in response to the third amount, a third number of time slots are allocated in each frame for transmitting third data to the first endpoint, such that the third number is variable from each frame to the next frame in the sequence, and such that the first and third numbers of time slots are not consecutive.

There is also provided in accordance with a preferred embodiment of the present invention apparatus for downstream communication in a passive optical network by time division multiplexing a sequence of frames, including:

a passive optical distribution structure adapted to receive and transmit data;

a plurality of receiving Optical Network Terminals (ONTs), coupled to the fabric comprising first and second ONTs, adapted to receive data from the fabric; and

a central Optical Line Terminal (OLT) coupled to the fabric and adapted to receive data for transmission into the network, the data including at least a first amount of first data for transmission to a first ONT and a second amount of second data for transmission to a second ONT, such that the first and second amounts are variable from each frame to a next frame in the sequence, such that a first number of time slots are allocated in each frame to transmit the first data to the first ONT and a second number of time slots are allocated to transmit the second data to the second ONT, such that the first and second numbers are variable from each frame to the next frame in the sequence in response to changes in the first and second amounts of data, and the data is transmitted during the allocated time slots.

Preferably, the data comprises data sets communicated via respective different industry standard services.

Preferably, the plurality of ONTs comprises sets of ONTs operable under different wavelength groups, and the data comprises respective sets of data communicated between the OLT and the respective sets of ONTs via the different wavelength groups.

It is further preferred that the data comprises sets of data communicated via respective different channels, wherein each channel transmits data via a service coupled with the OLT and at least one ONT.

It is also preferred that each channel is allocated a respective bandwidth and that allocating the first and second numbers of time slots comprises allocating the first and second numbers of time slots in response to the bandwidth of each channel.

Preferably, allocating the respective bandwidth to each channel comprises changing the respective bandwidth to a different bandwidth in response to a request received by the OLT.

Preferably, the OLT comprises a memory and the first and second numbers of time slots are allocated in response to respective first and second data parameters stored in the memory.

It is further preferred that the sum of the first number and the second number of time slots is smaller than or equal to the bandwidth of each frame sequence.

Preferably, the data comprises one or more further quantities of data for transmission to a respective one or more further ONTs of the plurality of ONTs such that the one or more further quantities are variable from each frame to the next frame in the sequence, wherein the OLT is adapted to allocate the respective one or more further quantities of time slots to be transmitted to the one or more further ONTs respectively.

Preferably, the period of each frame sequence is substantially constant.

Preferably, each frame sequence comprises a header comprising respective window parameters for each of the plurality of ONTs, each window parameter comprising a time and a size of a window of upstream data that each of the plurality of ONTs is allowed to transmit to the OLT.

Preferably, the corresponding window parameters are allocated by said OLT, so that the windows do not collide at the OLT.

It is further preferred that the data comprises at least a third amount of third data for transmission to the first endpoint, such that the third amount is variable from each frame to the next frame in the sequence, and in response to the third amount, the OLT is adapted to allocate a third number of time slots in each frame for transmitting the third data to the first endpoint, such that the third number is variable from each frame to the next frame in the sequence, and such that the first and third numbers of time slots are not consecutive.

Further in accordance with a preferred embodiment of the present invention, there is provided a method for communicating between a transmission point of a network and an end point of the network by time division multiplexing of a sequence of frames, each frame being divided into a plurality of time slots, the method comprising:

receiving at the transmission point data for transmission to an endpoint, the data comprising at least a first amount of first data for transmission to the endpoint and a second amount of second data for transmission to the endpoint, such that the first and second amounts are variable from each frame to the next frame in the sequence;

allocating a first number of timeslots in each frame to transmit first data to the endpoint and a second number of timeslots to transmit second data to the endpoint in response to the first and second amounts, such that the first and second numbers are variable from each frame to a next frame in the sequence in response to changes in the first and second amounts of data; and

transmitting the data from the transmission point to the endpoint during the allocated time slot.

Preferably, between the transmission point and the end point, the first data is communicated via a first channel and the second data is communicated via a second channel, wherein the first data is communicated via a first service and the second data is communicated via a second service, the first and second services being coupled to the transmission point and the end point and being external to the network.

Preferably, the data comprises a third amount of the first data for transmission to the endpoint, such that the third amount is variable from each frame to the next frame in the sequence, and comprising, in response to the third amount, allocating a third number of time slots in each frame for transmitting the third amount to the endpoint, such that the third amount is variable from each frame to the next frame in the sequence, and such that the first and third numbers of time slots are not consecutive.

There is also provided in accordance with a preferred embodiment of the present invention apparatus for communicating in a network by time division multiplexing a sequence of frames, including:

a receiver coupled to a network for receiving data from the network; and

a transmitter coupled to the network and adapted to receive data for transmission into the network, the data including at least a first amount of first data for transmission to a receiver and a second amount of second data for transmission to the receiver, such that the first and second amounts are variable from each frame to a next frame in the sequence, to allocate a first number of time slots in each frame to transmit the first data to the receiver and to allocate a second number of time slots to transmit the second data to the receiver, such that the first and second numbers are variable from each frame to the next frame in the sequence in response to changes in the first and second amounts of data, and to transmit the data during the allocated time slots.

Preferably, between the transmitter and the receiver, the first data is communicated via a first channel and the second data is communicated via a second channel, wherein the first data is communicated via a first service and the second data is communicated via a second service, the first and second services being coupled to the transmitter and the receiver and being external to the network.

Preferably, the data comprises a third amount of the first data for transmission to the receiver, such that the third amount is variable from each frame to the next frame in the sequence, and comprising, in response to the third amount, allocating a third number of time slots in each frame for transmitting the third amount to the receiver, such that the third number is variable from each frame to the next frame in the sequence, and such that the first and third numbers of time slots are not consecutive.

The present invention may be more completely understood in consideration of the following detailed description of preferred embodiments of the invention in connection with the accompanying drawings, in which:

drawings

Fig. 1 is a schematic diagram of a passive optical network in accordance with a preferred embodiment of the present invention;

fig. 2 is a schematic timing diagram illustrating the overall structure of a downstream frame and a virtual upstream frame transmitted in the network of fig. 1 in accordance with a preferred embodiment of the present invention;

fig. 3 is a diagram illustrating a structure of a downlink frame header in accordance with a preferred embodiment of the present invention;

FIG. 4 is a flow chart showing how the header of FIG. 3 is computed in accordance with a preferred embodiment of the present invention;

FIG. 5 is a diagram illustrating a payload section of a downstream frame in accordance with a preferred embodiment of the present invention;

FIG. 6 is a flow chart showing how payload segments are calculated in accordance with a preferred embodiment of the present invention;

FIG. 7 is a diagram illustrating details of an upstream window of the network of FIG. 1 in accordance with a preferred embodiment of the present invention; and

fig. 8 is a flowchart illustrating the steps involved in implementing channel bandwidth variation in accordance with a preferred embodiment of the present invention.

Detailed Description

Referring now to fig. 1, fig. 1 is a schematic diagram of a Passive Optical Network (PON)20 in accordance with a preferred embodiment of the present invention. The PON 20 includes an Optical Line Terminal (OLT)22 at the front end of a passive optical distribution structure 24 that serves as the central transmission point and overall control equipment for the PON 20. The fabric 24 terminates on its downstream side by generally similar Optical Network Terminals (ONTs) 26A, 26B, 26c. Hereinafter, ONT 26A, ONT 26B, ONT26℃. is also referred to herein generally as ONT26 and as ONT a, ONT B, ONT C.. respectively. The OLT22 and ONTs 26 operate as data modems, so that the OLT22 is coupled on their upstream side to industry standard data transmission services, such as an ethernet line 28, a video line 29, and a Constant Bit Rate (CBR) line 30, and the ONTs 26 are coupled on their respective downstream sides to data lines 32, which provide corresponding or low rate services to downstream clients of the PON 20. The PON 20 transmits data between the OLT22 and the ONTs 26 in downstream frames and upstream "virtual" frames, which are sent in a full-duplex manner. Full duplex methods, such as transmitting data using different transmit and receive wavelengths, are well known in the optical network arts. The downlink and uplink frame formats are described in more detail below.

Downstream frames from the OLT22 are sent into the fabric 24 in a substantially continuous sequence of constant period frames. Most preferably, the downlink frame has a 125 μ s periodicity and is implemented to transmit data at a rate of approximately 2.5Gb/s, although other periodicities and rates may be used. The fabric 24 passively separates the downstream transmissions so that all ONTs 26 receive the frames in a common broadcast fashion. In the upstream direction, the separate transmissions from the plurality of ONTs 26 are sent as windows incorporating the virtual frame so that they do not collide when they reach the OLT 22. The virtual upstream frame is implemented to have substantially the same period as the downstream frame. Preferably, the upstream data transmission is sent at a rate approximately equal to the downstream rate, although other upstream rates may be used.

The OLT22 includes a Control and Logic Unit (CLU)36 for controlling the overall operation of the PON 20 and the operation of the individual elements of the OLT via management software 33 contained in a memory 44. The OLT22 is coupled to a network monitor 37 and a keyboard 35 that enable an operator of the PON 20 to track network behavior and effect changes to the network via software 33. The OLT22 comprises a switch 34 for switching between services coupled to the OLT in accordance with control signals received from the CLU 36. The CLU also controls the operation of the data entry (from the upstream direction) FIFO logic 38, data exit (to the upstream direction) FIFO logic 40, transmit framer 31, receive framer 23, and FIFO parameters table 21. The function and operation of these FIFOs, framers, tables 21, and the function and operation of the management software 33 will be described in more detail below.

The OLT22 also includes an optical interface 42, which is controlled by the CLU 36. The interface 42 converts the data from the transmit framer 31 to an optical format, most preferably by modulating the laser light contained in the interface and transmitting the laser output into the structure 24. The interface 42 also receives optical signals from the fabric 24 and converts those signals into data to be sent to the receive framer 23 later, where it is decomposed before being transmitted upstream via the FIFO 40 and the switch 34.

Preferably, the interface 42 is capable of transmitting and receiving its optical output substantially simultaneously in groups of multiple discrete wavelengths [ λ 1], [ λ 2], [ λ 3 ]. to effectively increase the capacity of the PON 20 by the number of wavelength groups used. Each wavelength group includes a first wavelength at which the OLT22 sends downstream data and a second wavelength at which the OLT receives upstream data. The PON 20 comprises a set of ONTs 26, each ONT26 in a particular set operating in one of the wavelength groups by receiving a first wavelength and transmitting a second wavelength. Alternatively, the interface 42 transmits and receives in one wavelength group so that all ONTs 26 are included in one set. The format of the frames transmitted in the multi-wavelength system is substantially the same as in the single-wavelength system, and thus for the sake of clarity, it is assumed that the PON 20 operates in one wavelength group unless otherwise specified below.

Each ONT26 includes a CLU46, an ingress FIFO 48, an egress FIFO50, a transmit framer 41, a receive framer 56, a FIFO parameter table 51, an optical interface 52, a switch 64, and a memory 54, which operate in a manner similar to CLU36, ingress FIFO38, egress FIFO 40, transmit framer 31, receive framer 23, table 21, optical interface 42, switch 34, and memory 44, respectively. When multiple discrete sets of wavelengths [ λ 1], [ λ 2], [ λ 3],. are used in PON 20, each optical interface 52 is implemented to transmit and receive data in one of the discrete sets, but not in response to the other set of wavelengths, most preferably by filters in the optical interface.

Communications in the PON 20 are classified into channels, each of which transmits data for a particular service between the OLT22 and a particular ONT26 or set of ONTs, so that there is a one-to-one mapping between the services supported by the PON and the channels used for transmission. OLT22 and each ONT26 maintain a primary and secondary table of channel parameters. The OLT22 includes a primary downstream channel table 43 in memory 44 for mapping channels to their assigned downstream bandwidths and channel labels and a secondary downstream table 45 for use when making adjustments to the channel downstream bandwidth. OLT22 also includes a primary and secondary pointer tables 47 and 49 for tracking and updating ONT26 window parameters, respectively. The OLT22 also includes a table 25 for mapping the ONTs 26 and corresponding channel labels. The functions of tables 25, 43, 45, 47, and 49 will be described in more detail below.

Each ONT26 includes a primary upstream channel table 53 in a respective memory 54, each primary table mapping the channels of a particular ONT26 to their allocated upstream bandwidth and channel label. Each ONT26 also includes a corresponding secondary upstream channel table 55, which is used in making adjustments to the upstream bandwidth of the channel, and a table 59 of downstream channel labels. A more detailed description of the operation of these tables will be given below.

In the PON 20, channels are defined for services coupled to the PON. Most preferably this provision is made by the operator of the PON via a keyboard 35 and monitor 37, the operator also allocating resources to the channel, such as upstream and downstream bandwidth.

Most preferably, the channels are classified according to their specification as Constant Bit Rate (CBR) channels or Packet (PB) based channels. For example, a channel for data originating from the ethernet line 28 is generally specified as a PB channel. When a channel is specified, parameters including a channel label and upstream and downstream bandwidths allocated to the channel are entered into the respective upstream and downstream master tables. Most preferably, the bandwidth is allocated in 4 byte slots, although any other convenient slot size may be used. Most preferably, the bandwidth allocated to a particular channel is greater than the input data rate. For example, for a CBR channel on line 30 (where the frequency is 1.544Mbit/s), the necessary bandwidth for a 125 μ s frame is 193 bits, which is accommodated by setting the channel bandwidth to 74 byte time slots. Details of the frame characteristics for the network 20, as well as the bandwidth and adjustments to the bandwidth allocated to particular channels, including the PB channel, are given below.

The ingress FIFO38 acts as an initial buffer for downstream data to the OLT22 and most preferably comprises a separate memory according to the type of service coupled to the OLT 22. Data from the CBR channel is written into the FIFO38 at the clock rate served by the CBR. Most preferably, only valid packets, i.e., packets without invalid bytes, are written from the PB channel into the FIFO 38. After receipt from lines 28, 29 and 30 and after transmission via switch 34, the data is written into FIFO 38. The channel parameters written into the FIFO38 are written into the FIFO parameter table 21 as appropriate. For example, for a PB channel, it is preferable that table 21 includes the maximum burst size allocated (MBS), the maximum burst rate allocated (MBR) and the Guaranteed Bit Rate (GBR) for the channel; for CBR signaling, it is preferred that table 21 include the GBR of the channel.

During the read from FIFO38, CLU36 reads the allocated downstream bandwidth of the channel from main table 43 and the data from table 21. Up to the allocated bandwidth, data subject to any limitation by the parameters in table 21 is transferred by CLU36 from FIFO38 to transmit framer 31. Framer 31 is used by CLU36 to assemble the data before transmission to fabric 24. The transmission mode of the downlink data will be described in more detail below.

In ONT26, the ingress FIFO 48 and the transmit framer 41 generally operate with reference to FIFO38 and framer 31, respectively. Upon reading from FIFO 48, each CLU46 reads the channel upstream bandwidth from the corresponding main table 53 and the channel parameters from table 51. For each ONT26 data subject to the limitations of table 21, the data is transferred from FIFO 48 by CLU46 to transmit framer 41, which is used to assemble the data before transmission into fabric 24. The transmission of the uplink data will be described in more detail below.

Fig. 2 is a schematic timing diagram showing the overall structure of a downstream frame 70 and a virtual upstream frame 72 in accordance with a preferred embodiment of the present invention. Each downstream frame 70 is transmitted from the OLT22 with a substantially equal period, hereinafter referred to as a network period, most preferably set to 125 mus. Each downstream frame 70 includes a header 74 that may be used, at least in part, as a means for receiving an ONT to identify the start of the frame, among other factors, and thus serves as a start identifier throughout the frame timing of the PON 20. Each downstream frame 70 also includes a payload section 76 in which data transmitted from the OLT22 (from the services sent on lines 28, 29 and 30) to the ONTs 26 is input. The downlink frames 70 are transmitted in a substantially continuous manner with no idle time between adjacent frames.

Each ONT26 is capable of transmitting upstream data during each network cycle defined by the corresponding adjacent downstream header 74. For each period, OLT22 assigns to each ONT26 a respective time window 78 in which it can send upstream data and is divided among the windows according to the channel. Each window 78 includes a start time associated with the start identifier defined by the frame header 74 and a length of time during which a particular ONT is allowed to be transmitted. For example, windows 78 of adjacent virtual frames 72 for ONT A, ONT B, and ONT N are shown in FIG. 2. As described in more detail below, the window start time and length are allocated within each virtual frame 72 such that the windows 78 do not overlap, thereby ensuring that data from different ONTs at the OLT22 do not collide. The allocation is determined by the CLU36 and communicated to the ONTs 26 in the downstream frame 70. It should be understood that the start time and length of the window 78 is dynamic and may vary from frame to frame. In addition to communicating upstream data for the corresponding service in the channel, window 78 may also communicate other data, as described above. The data in such windows includes bandwidth change requests and distance information data, which will be described below, as well as other management and control windows.

Fig. 3 is a diagram illustrating the structure of the downlink frame header 74 according to the preferred embodiment of the present invention. The header 74 includes a frame identification section 80, which in turn includes a Frame Alignment Signal (FAS) section 84, a wavelength identification section 83, a plurality of ONT sections 85, and a Bit Interleaved Parity (BIP) section 86. Preferably, the FAS segment 84 comprises eight bytes and is used by the ONT as an identifier of the start of frame. When the PON 20 operates with more than one wavelength, section 83 identifies which of the wavelength groups [ λ 1], [ λ 2], [ λ 3],. the frame is using, and section 85 indicates the number of ONTs operating in that wavelength group. When a single wavelength group is used for the PON 20, the segment 85 indicates the number of ONTs operable within the network. The BIP segment 80 is used as an error monitoring function with an even check computed on all bits immediately preceding frames, and preferably includes a BIP-8 code function, as is well known in the art.

The header 74 also includes one or more ONT header sections 82, each section 82 being dedicated to a particular ONT 26. The number of header sections 82 corresponds to the number of ONTs 26 that communicate with the frame 70. Each ONT header section 82 includes a one byte ONT identification section 88, and a window pointer section 90. The pointer segment 90, which is preferably three bytes long, includes information to be used by a particular ONT in sending its data upstream. Pointer segment 90 includes a start window transmission time 91 (calculated by receipt of FAS segment 84), which is the time that the particular ONT begins transmitting its data within the corresponding window 78. The segment 90 also includes a maximum length 93 of the window 78 that the ONT is allowed to transmit. The length 93 corresponds to the total upstream bandwidth allocated to a particular ONT26 and is calculated by the management software 33 from the aggregate upstream channel bandwidth requests received from all ONTs 26.

The grant segment 92, which is preferably 4 bits long, includes a control code used by the OLT22 to control upstream transmissions from each ONT. Preferably, the control code comprises code translated by each receiving ONT to describe data that each particular ONT should or should not transmit. For example, the permission segment 92 instructs a particular ONT to send its window 78 to stop transmission, or to send one of a plurality of management or control types of data windows.

The ONT management and control channel segment 94 may be used as a protection check for segments 88, 90, and 92, among other factors. Segment 94 includes a Cyclic Redundancy Check (CRC) segment 95, a remote monitoring segment 96, a Data Communication Channel (DCC) segment 97, and a Fast Communication Channel (FCC) segment 98. CRC section 95 is used to protect section 88, section 90 (including time 91 and length 93), and section 92. The ONT identified in section 88 compares the CRC received in section 95 to the expected values for sections 88, 90 and 92. If an error is detected by the comparison, then most preferably the ONT stops upstream transmission until the frame is received with the correct CRC in segment 95. The remote monitoring segment 96 includes a Remote Defect Indication (RDI) and a remote error indication (RBI). When the OLT22 detects a defect, such as the absence or loss of a window 78 from an individual ONT, or if there are too many errors in the received window, an RDI is sent from the OLT22 to a particular ONT. The REI includes a count of a number of errors detected by the OLT22 in the last data window received by the OLT.

DCC section 97 and FCC section 98 include areas that enable OLT22 to send management information specific to ONT26, and/or any other management-related information. For example, changes in the downstream bandwidth of a particular channel handled by the ONT26 may be transmitted via the DCC segment 97. Preferably, the decoding of the section 97 is implemented by software in a particular ONT 26. Most preferably, FCC section 98 is implemented by a combination of hardware and software at the physical level of the ONT, such that section 98 may be used to communicate management and/or control messages to ONT26 at a faster rate than DCC section 97.

Fig. 4 is a flow chart showing how the CLU36 calculates the header 74 using the management software 33 in accordance with the preferred embodiment of the present invention. Most preferably, the steps of the flowchart are implemented by the CLU36 before each downstream frame 70 is sent from the OLT. In a first step, the CLU36 determines how many ONTs 26 are active in the PON 20. Methods for performing such a determination are known in the art and include recording which ONTs 26 have sent data to the OLT22 during a predetermined time period before performing this step. Most preferably, the method used includes a method for recording when a new ONT26 enters the stream, and when the ONT26 leaves the stream.

In a second step, the CLU36 calculates the maximum window size of each window 78 in terms of time slots for the respective active ONT26 determined in the first step. CLU36 sets the total number of maximum sizes to be less than the downstream bandwidth capability. Within this limit, the window size for each ONT26 is set according to the data parameters of each active channel communicated upstream by a particular ONT 26. Preferably, the data parameters for each channel include whether the channel is a CBR or PB channel, the priority assigned by the operator of PON 20, the amount of data for a particular channel in the corresponding ingress FIFO 48, and the bandwidth allocated to the channel in the main table 53 of ONTs 26. Most preferably, the data parameters for each PB channel also include an upstream Guaranteed Bit Rate (GBR) assigned by the network operator and an upstream Maximum Burst Rate (MBR) (the time at which the bandwidth change is effected, either by specifying the channel or, as described below, the time at which the bandwidth change is effected). Most preferably, the number of timeslots allocated to each PB channel exceeds the uplink GBR. Most preferably, the number of time slots allocated to each CBR channel exceeds the worst-case possible uplink frequency of the channel.

Most preferably, at least some of the data parameters are transmitted in management and control packets transmitted by a particular ONT26, and preferably at least some are stored in the FIFO parameter table 21. Most preferably, the management and control packets are sent by ONT26 in one or more preceding windows 78 in response to a request generated by software 33 before OLT22 calculates the window size for ONT 26.

In a third step, CLU36 preferably places windows 78 calculated in step two consecutively in a "virtual" frame, and determines for each window 78 an initial start time relative to the frame start time, corresponding to the start of header 74. As described below, a distance time and an edge time are added to each initial start time to generate a start window transmission time to be used for each window 78.

Each ONT26 is some distance from the OLT22 that produces a delay in the transmission of signals from a particular ONT26 to the OLT22 that is proportional to the distance due to the finite transmission rate of the optical signals. The distance of each ONT26 from the OLT22 is compensated by adding a distance time to the initial time window start time of each ONT. The range time for each ONT26 is determined by one of the ranging methods known in the art of passive optical networks, such as by sending a ranging signal from the OLT22 into the PON 20 and waiting for a corresponding response to be received from an active ONT 26. Preferably, the ranging signal is transmitted when the PON 20 is initialized. Alternatively or additionally, ranging signals are sent when operating network 20, such as when OLT22 determines that a new ONT26 has entered the stream, or when there has been a valid change in an operating parameter of an existing ONT 26. Once ranging signal responses have been received from each ONT26 operating in the PON 20, the OLT22 determines the range time for each ONT 26.

The CLU36 adds an edge time and an instance time to the initial start time to ensure that there is some interval of instants to reach the window 78 of the OLT 22. Due to the following factors: such as differences and/or drift in the clocks of each ONT26, physical changes in the components of the PON 20 over time and errors in the ranging process, which also allows for errors in the arrival time of the OLT 22.

In a final step, CLU36 inserts the calculated start window transmission time, maximum window size, and CRC section 95 into pointer section 90 of each ONT header section 82. The time and size of each window is also stored in pointer table 47, according to the particular ONT 26.

Fig. 5 is a schematic diagram illustrating a payload section 76 of a downstream frame 70 in accordance with a preferred embodiment of the present invention. The payload section 76 includes data sent from the OLT22 to the plurality of ONTs 26. The data is incorporated into section 76 on a variable time multiplexed basis, so for each frame 70, OLT22 allocates one or more data channels 100 for each ONT to which the data is to be sent. The size of each data channel will be calculated by the CLU36 using the management software 33 in accordance with a predetermined standard, as will be described below with reference to fig. 6, and the size of the corresponding data channel can vary from frame to frame for each ONT 26.

The data channel is made up of a set of time slots. Each data channel is identified by a channel tag (described below) that is read from table 43 and conveys data belonging to a single channel. Most preferably, each data channel 100 in the payload 76 is contiguous with an adjacent data channel 100. Alternatively, adjacent data channels 100 are separated by one or more "spare bits". Each data channel 100 includes a channel overhead section (overhead section)102 and a channel payload section 104. The channel overhead segments 102 are substantially similar in size and construction. The channel payload sections 104 typically differ in size from one another, from frame to frame, and the size of each channel is calculated on an ongoing basis.

Each channel overhead section 102 includes a channel tag field 106, a length field 108, a management and control field 110, and a CRC section for guarding preceding data. The channel label field 106 includes a unique label that is specified by the CLU36 when the channel is provisioned and is stored in the table 43 of the OLT22 and the corresponding table 59 of the associated ONT 26. The channels, their lengths, and their labels are tracked in the OLT22 and each ONT26 as described below.

The length field 108 gives the size of the payload section 104 of the associated data channel 100 in terms of time slots. Most preferably, field 108 is divided into two segments, a 2-bit multiplier segment, and a 10-bit length segment. The magnitude of the payload is calculated by multiplying the value associated with the 2-bit value by the 10-bit value. Preferably, the management and control field 110 is substantially similar in form and function to the management and control segment 94 (FIG. 3). Most preferably, field 110 includes subfields that enable OLT22 to set channel priority and/or generate channel alarms, such as when an incorrect CRC checksum has been previously received in a particular channel. Preferably, the field 110 also enables the OLT22 to communicate channel management information to a particular ONT 26.

Fig. 6 is a flow chart illustrating construction of payload section 76 using management software 33 in accordance with a preferred embodiment of the present invention. The steps of the flowchart are performed before each downstream frame 70 is sent from the OLT 22.

In a first step, CLU36 checks the entry FIFO38 for data to be merged into segment 76.

In a second step, for each channel of data in FIFO38, CLU36 determines the allocated downstream bandwidth of the channel and the channel tag from main table 43. For each channel, CLU36 also determines the data parameters in table 21.

In a third step the allocated bandwidth up to the channel is read from the FIFO38, as well as data for each channel subject to any restrictions generated by the parameters in table 21.

In a fourth step, a channel overhead section 102 incorporating the channel tag and length is constructed and the data read from the FIFO is incorporated into a channel payload section 104 to form the data channel 100 for the particular channel.

Steps two through four are repeated for each channel of data in the FIFO38 and in a final step, the data channels 100 generated for each channel are concatenated to form the downstream payload section 76. Most preferably, steps two through four and the final step are implemented by CLU36 using send framer 31 when constructing segment 76.

It should be appreciated that the downstream data frames 70 are sent in a broadcast manner, and thus a particular ONT26 receives data that is not directed to that ONT. Most preferably, the data sent in payload section 104 is encrypted so that only the ONT to which the data is directed can decrypt the data. Such methods of encryption and decryption are well known in the art and include "churning" the data. If churning is used, it is most preferred that a churning key is generated by each ONT 26. And sends it to OLT22 in management and control upstream window 78, OLT22 then churns the channel payload data before it enters payload section 104. Unstirred (De-churning) is then performed at the ONT using the churning key. Most preferably, new mix keys are sent from each ONT26 at regular intervals of less than 1 second.

It should be appreciated that the structure of the downlink payload section 76 allows the bandwidth of the channel used for transmission to be implemented for each downlink frame 70 to be varied. Most preferably, this change is accomplished using secondary table 45, as will be described in greater detail with reference to FIG. 8.

The downstream payload section 76 is constructed on a time multiplexed "per channel" basis, with each data channel 100 being defined by its channel header 102 (fig. 5). By comparing label 106 in header 102 with the channel labels in table 59 of each ONT26, each particular ONT26 recovers the data channel directed to it from section 76. It should be understood that the data channels directed to a particular ONT26 are not necessarily contiguous in segment 76, as each ONT26 checks the channel label. Thus, any particular ONT26 may receive its time-multiplexed data in segments, which are per channel. It will be appreciated that allowing data to be sent to a particular ONT to be segmented significantly increases the "packetization" efficiency of segment 76.

Fig. 7 is a schematic diagram illustrating details of upstream window 78 of ONT 26B in accordance with a preferred embodiment of the present invention. The data in upstream window 78 is generated and transmitted by ONT 26B, although it is clear that the following description is applicable to the transmission and generation of data in upstream window 78 from any ONT26, taking into account differences in detail. The window 78 is implemented by the ONT 26B so that when the window is received at the OLT22, the window does not overlap any other windows 78 transmitted by other ONTs 26, as schematically illustrated in FIG. 2. Window 78 is only transmitted from ONT 26B after being received by the ONT of ONT 26B overhead section 82 (FIG. 3). As described with reference to FIG. 3, segment 82 includes a pointer segment 90, pointer segment 90 having a start window transmission time 91 and a maximum window length 93.

Window 78 includes a generic ONT 26B overhead section 134 followed by alternating channel overhead and channel payload sections. For example, assume that window 78 for ONT 26B includes two channels, each having an overhead section 136 and 140, and a payload section 138 and 142. The overhead section 134 includes a synchronization field 146, an identification field 148, a window type field 149, a CRC field 156, a BIP field 150, a DCC/FCC field 152, and an RDI/REI field 154. Fields 146 and 148 enable OLT22 to identify ONT 26B as the transmitting ONT and also enable the OLT to synchronize a burst mode receiver contained within the OLT. The field 149 indicates the window composition, such as a data window. The CRC field 156 provides a checksum for protecting the data transmitted in the synch field 146, identification field 148, and window type field 149. BIP field 150 is implemented and operates in a generally similar manner as BIP segment 80 (fig. 3). DCC/FCC field 152 is implemented and operates in a manner generally similar to segments 97 and 98 to provide a field for general management and control messages for ONT 26B. The RDI/REI field 154 is implemented and operates in a generally similar manner as the segment 96, which acts as a remote monitoring field.

Each channel overhead section 136 and 140 includes a channel label field 158 and a channel length field 160. Field 158 is derived from channel label table 53 in ONT 26B. Most preferably, the field 160 is substantially similar in form to the length field 108 described above with reference to FIG. 5. Preferably, each channel overhead section also includes a status field 162 for communicating alarms regarding a particular channel, e.g., to OLT22, and a CRC field 164 that includes checksum protection fields 158, 160, and 162. Each channel overhead segment is followed by its channel payload segment, having a length defined by its corresponding channel overhead segment 160.

The length of each channel payload section is determined by CLU46 of ONT 26B according to the following overall constraint: the length of the window 78 is less than or equal to the maximum window length 93 received in the pointer 90. Within this limitation, the CLU46 constructs a channel payload segment by reading the special channel data, data parameters and the respective allocated bandwidth for each channel from the ingress FIFO 48, FIFO parameter table 51 and main table 53. After constructing window 78, ONT 26B transmits the window upstream at a start window transmission time 91 after receiving header 74.

Fig. 8 is a flow chart showing the steps involved in implementing channel bandwidth variation in accordance with a preferred embodiment of the present invention. Typically, when a channel is defined, the channel is assigned upstream and downstream bandwidths in accordance with the type of service providing the channel and/or in accordance with a Service Level Agreement (SLA) between an operator of PON 20 and a network user. Each channel tag and corresponding downstream bandwidth is written into the main table 43 of the OLT 22. The master table 53 for each ONT26 contains the channel label and upstream bandwidth of the channel transmitted via the particular ONT 26. The flow chart of fig. 8 is implemented by the CLU36 using the management software 33.

In a first step, a request for a change in channel bandwidth is received by the CLU 36. Preferably, the request is initiated by a user of the channel and typically includes a request to increase the upstream and/or downstream bandwidth of the channel. Most preferably, the request is made by a network operator via a keyboard 35.

In a second step, CLU36 detects whether sufficient free bandwidth is available to accommodate any requested increase by evaluating the total number of time slots that have been allocated to downstream frames 70 and/or to upstream virtual frames 72. If there is sufficient free bandwidth available, then the request to change bandwidth continues; if insufficient bandwidth is available, the process stops.

In a third step, the CLU36 reallocates time slots for the channel. If the new bandwidth allocation is for downstream bandwidth, the reallocated time slots may be in existing data channel 100 locations (fig. 5). Alternatively, by generating a second data channel 100 with a tag 106 corresponding to the channel, more bandwidth can be allocated to a particular channel. The two data channels may reside anywhere in segment 76, and not necessarily contiguous, thereby increasing the flexibility with which segment 75 may be set, as well as the packing efficiency of the segments.

If the new bandwidth allocation is for upstream bandwidth, the size and position of the window 78 is adjusted as needed. For example, referring to fig. 7, increasing the bandwidth of channel 1 may be accomplished by adding a second segment of channel 1 (the segment comprising the channel 1 overhead and payload segments) in region 144, maintaining segment 93 fixed. The two segments of channel 1 need not be contiguous and so the advantages described for the downlink frame apply equally.

If a greater increase in upstream bandwidth is requested, segment 93 may have to be increased in size and/or the positions of other windows 78 may have to be adjusted within virtual frame 72. The change in size and/or position of the window 78 is communicated to the ONT26 via an appropriate segment 90 of the downstream frame 70.

In a fourth step, the channel change is sent via downstream frame 70 to the appropriate ONT26, i.e. those ONTs 26 affected by the implemented change. The changes may be transmitted via segment 94 (fig. 3) and/or segment 110 (fig. 5) of frame 70 and incorporated into the corresponding secondary table 55.

In a final step, in the OLT22, the primary table 43 is replaced by the table 45 and the primary tables 53 are replaced by their respective secondary tables 55. The FIFO parameter tables 21 and 51 are also updated as necessary. Operation of the PON 20 then continues with the updated bandwidth.

It should be understood that the generally same procedure as described above applies to de-provisioning (de-provisioning) of new channels and existing channels.

It should also be understood that the preferred embodiments of the present invention may be implemented in communication networks, including networks that include a transmitter and a receiver in addition to passive optical networks such as the PON 20. All such networks are included within the scope of the present invention. Furthermore, it should be understood that the data transmitted by the preferred embodiment of the present invention is not specific to any one protocol or service, but can be transmitted into and from the network in substantially any protocol or service.

It will thus be appreciated that the preferred embodiments described above are cited by way of example, and that the present invention is not limited to what has been particularly shown and described hereinabove. In addition, the scope of the present invention includes both combinations and subcombinations of the various features described hereinabove, as well as variations and modifications thereof which would occur to persons skilled in the art upon reading the foregoing description and which are not disclosed in the prior art.

Claims (33)

1. A method of downstream communication from a central transmission point (22) to a plurality of receiving endpoints (26A, 26B, 26℃.) by time division multiplexing of a sequence of frames, wherein each frame is divided into a plurality of time slots, the method comprising:

receiving at the central transmission point data for transmission to the end points, the data comprising at least a first amount of first data for transmission to a first end point (26A) of the plurality of end points and a second amount of second data for transmission to a second end point (26B) of the plurality of end points, such that the first and second amounts are variable from each frame to the next frame in the sequence;

allocating a first number of timeslots in each frame to transmit first data to the first endpoint and a second number of timeslots to transmit second data to the second endpoint in response to the first and second amounts, such that the first and second numbers are variable from each frame to the next frame in the sequence in response to changes in the first and second amounts of data; and

the data is transmitted from the central transmission point to the end points during the allocated time slots.

2. A method as recited in claim 1, wherein the central transmission point comprises an Optical Line Terminal (OLT) and the end points comprise Optical Network Terminals (ONTs), wherein the OLT and ONTs are operable as transceivers in a passive optical network.

3. The method of claim 1 or 2, wherein the data comprises a data set communicated via respective different industry standard services.

4. The method of any of claims 1-3, wherein the plurality of endpoints comprises sets of endpoints operable at different wavelength groups, and wherein the data comprises respective sets of data communicated between a central transmission point and respective sets of endpoints via different wavelength groups.

5. A method according to any of claims 1-4, wherein the first and second number of time slots are allocated in response to respective first and second data parameters stored in a memory comprised in the central transmission point.

6. The method of any of claims 1-5, wherein a sum of the first number and the second number of slots is less than or equal to a bandwidth of each frame sequence.

7. A method according to any of claims 1-6, wherein the data comprises one or more further quantities of data for transmission to one or more further endpoints of the plurality of endpoints, such that the one or more further quantities are variable from each frame to the next frame in the sequence, and comprising allocating respective one or more further numbers of time slots to be transmitted to the one or more further endpoints respectively.

8. The method of any of claims 1-7, wherein a period of each frame sequence is substantially constant.

9. A method according to any of claims 1-8, wherein the data comprises at least a third amount of third data for transmission to the first endpoint, such that the third amount is variable from each frame to the next in the sequence, and comprising, in response to the third amount, allocating a third number of time slots in each frame to transmit the third data to the first endpoint, such that the third number is variable from each frame to the next in the sequence, and such that the first and third numbers of time slots are not consecutive.

10. The method of any of claims 1-9, wherein the data comprises a data set communicated via respective different channels, wherein each channel transmits data via a service coupled to a central transmission point and at least one endpoint.

11. The method of claim 10, comprising allocating a respective bandwidth to each channel, and wherein allocating the first and second numbers of time slots comprises allocating the first and second numbers of time slots in response to the bandwidth of each channel.

12. The method of claim 11, wherein allocating a respective bandwidth to each channel comprises changing the respective bandwidth to a different bandwidth in response to a request received by the central transmission point.

13. The method of claim 1, wherein each frame sequence comprises a header comprising a respective window parameter for each of a plurality of endpoints, each window parameter comprising a time and a size of a window of upstream data each of the plurality of endpoints is allowed to transmit to the central transmission point.

14. The method of claim 12, wherein the respective window parameters are assigned by the central transmission point so that windows do not collide at the central transmission point.

15. An apparatus for downstream communication in a passive optical network by time division multiplexing of a sequence of frames, comprising:

a passive optical distribution structure (24) adapted to receive and communicate data;

a plurality of receiving Optical Network Terminals (ONTs) (26A, 26B, 26℃.) coupled to the fabric, including first and second ONTs (26A, 26B) adapted to receive data from the fabric; and

a central Optical Line Terminal (OLT) (22) coupled to the fabric and adapted to receive data for transmission into the network, the data including at least a first amount of first data for transmission to a first ONT and a second amount of second data for transmission to a second ONT, such that the first and second amounts are variable from each frame to a next frame in the sequence, to allocate a first number of time slots in each frame to transmit the first data to the first ONT and a second number of time slots to transmit the second data to the second ONT, such that the first and second numbers are variable from each frame to the next frame in the sequence in response to changes in the first and second amounts of data, and to transmit the data during the allocated time slots.

16. The apparatus of claim 15, wherein the data comprises a data set delivered via respective different industry standard services.

17. The apparatus of claim 15 or 16, wherein the plurality of ONTs comprises sets of ONTs operable at different wavelength groups, and wherein the data comprises respective sets of data communicated between the OLT and the respective sets of ONTs via the different wavelength groups.

18. The apparatus of any of claims 15-17, wherein the OLT comprises a memory, and wherein the first and second numbers of time slots are allocated in response to respective first and second data parameters stored in the memory.

19. The apparatus of any of claims 15-18, wherein a sum of the first number and the second number of time slots is less than or equal to a bandwidth of each frame sequence.

20. An apparatus according to any of claims 15-19, wherein the data comprises one or more further quantities of data for transmission to a respective one or more further ONTs of the plurality of ONTs such that the one or more further quantities are variable from each frame to a next frame in the sequence, wherein the OLT is adapted to allocate the respective one or more further numbers of time slots to be transmitted to the one or more further ONTs respectively.

21. The apparatus of any of claims 15-20, wherein a period of each frame sequence is substantially constant.

22. The apparatus of any of claims 15-21, wherein the data comprises a third amount of third data for transmission to the first endpoint, such that the third amount is variable from each frame to a next frame in the sequence, and wherein the OLT is adapted to allocate a third number of time slots in each frame to transmit the third data to the first endpoint, such that the third number is variable from each frame to the next frame in the sequence, and such that the first and third numbers of time slots are not consecutive, in response to the third amount.

23. An apparatus according to any of claims 15-22, wherein the data comprises a data set communicated via respective different channels, wherein each channel transmits data via a service coupled to the OLT and at least one ONT.

24. The apparatus of claim 23, wherein each channel is allocated a respective bandwidth, and wherein allocating the first and second numbers of time slots comprises allocating the first and second numbers of time slots in response to the bandwidth of each channel.

25. The apparatus of claim 24, wherein allocating the respective bandwidth to each channel comprises changing the respective bandwidth to a different bandwidth in response to a request received by the OLT.

26. The apparatus of claim 15, wherein each frame sequence comprises a header comprising a respective window parameter for each of a plurality of ONTs, each window parameter comprising a time and a size of a window of upstream data that each of the plurality of ONTs is allowed to transmit to the OLT.

27. The apparatus of claim 26, wherein respective window parameters are assigned by the OLT such that windows do not collide at the OLT.

28. A method of communicating between a transmission point (22) of a network (24) and an endpoint (26A) of the network by time-division multiplexing a sequence of frames, wherein each frame is divided into a plurality of time slots, the method comprising:

receiving at the transmission point data for transmission to the endpoint, the data comprising at least a first amount of first data for transmission to the endpoint and a second amount of second data for transmission to the endpoint, such that the first and second amounts are variable from each frame to the next frame in the sequence;

allocating a first number of timeslots in each frame to transmit first data to the endpoint and a second number of timeslots to transmit second data to the endpoint in response to the first and second amounts, such that the first and second numbers are variable from each frame to a next frame in the sequence in response to changes in the first and second amounts of data; and

transmitting the data from the transmission point to the endpoint during the allocated time slot.

29. The method of claim 28, wherein between the transmission point and the endpoint, the first data is communicated via a first channel and the second data is communicated via a second channel, wherein the first data is communicated via a first service and the second data is communicated via a second service, the first and second services being coupled to the transmission point and the endpoint and external to the network.

30. A method as claimed in claim 28 or 29, wherein the data comprises a third amount of first data for transmission to the endpoint, such that the third amount is variable from each frame to the next frame in the sequence, and comprising, in response to the third amount, allocating a third number of time slots in each frame to transmit the third amount to the endpoint, such that the third amount is variable from each frame to the next frame in the sequence, and such that the first and third numbers of time slots are not consecutive.

31. A device for communicating in a network (24) by time division multiplexing of a sequence of frames, comprising:

a receiver (26A) coupled to the network for receiving data from the network; and

a transmitter (22) coupled to the network and adapted to receive data for transmission into the network, the data including at least a first amount of first data for transmission to the receiver and a second amount of second data for transmission to the receiver such that the first and second amounts are variable from each frame to a next frame in the sequence, to allocate a first number of time slots in each frame to transmit the first data to the receiver and a second number of time slots to transmit the second data to the receiver such that the first and second numbers are variable from each frame to the next frame in the sequence in response to changes in the first and second amounts of data, and to transmit the data during the allocated time slots.

32. The apparatus of claim 31, wherein between the transmitter and receiver, the first data is communicated via a first channel and the second data is communicated via a second channel, wherein the first data is communicated via a first service and the second data is communicated via a second service, the first and second services being coupled to the transmitter and the receiver and external to the network.

33. An apparatus according to claim 31 or 32, wherein the data comprises a third amount of the first data for transmission to the receiver, such that the third amount is variable from each frame to the next frame in the sequence, and comprising, in response to the third amount, allocating a third number of time slots in each frame to transmit the third amount to the receiver, such that the third amount is variable from each frame to the next frame in the sequence, and such that the first and third numbers of time slots are not consecutive.