Highly available PON system

Abstract

A system supporting high availability for passive optical networks.

Description

[Background technology]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 63 / 399,969, filed August 22, 2022.

[0002] The subject matter of this application relates to high availability for passive optical networking.

[0003] Passive optical networks (PONs) are often employed as access networks, or portions of larger communications networks. Communications networks typically have a high-capacity core portion over which data or other information associated with telephone calls, digital television, and Internet communications is carried significant distances. The core portion may have the capability to interact with other networks to complete the transmission of telephone calls, digital television, and Internet communications. In this manner, the core portion in combination with the passive optical network enables communications to and from subscribers (or other devices associated with subscribers, customers, businesses, or otherwise).

[0004] The access network of a communications network extends from the core portion of the network to individual subscribers, such as those associated with a particular place of residence (e.g., place of business). The access network may be wireless access, such as a cellular network, or fixed access, such as a passive optical network or a cable network.

[0005] Referring to FIG. 1, in a PON 10, a set of optical fibers and passive interconnection devices are used for most or all of the communications over a range of access networks. A set of one or more optical network terminals (ONTs) 11 are devices typically located at subscriber residences (e.g., or places of business). The term "ONT" includes what are also referred to as optical network units (ONUs). There may be any number of ONTs associated with a single optical splitter 12. As an example, 32 or 64 ONTs are often associated with a single network optical splitter 12. The optical splitter 12 is interconnected with each ONT 11 by a respective optical fiber 13, or otherwise by a respective fiber within a fiber optic cable. Selected ONTs may be removed and / or added to the access network associated with the optical splitter 12 as needed. There may also be multiple optical splitters 12 arranged in a cascaded configuration.

[0006] Optical fiber 13 interconnecting optical splitter 12 and ONT 11 acts as an access (or "drop") fiber. Optical splitter 12 is typically located within a street cabinet or other structure in which one or more optical splitters 12 are located, each serving a respective set of ONTs. In some cases, an ONT may serve multiple subscribers, such as subscribers in multiple residential units (e.g., an apartment building). In this way, a PON may be considered a point-to-multipoint topology, in which a single optical fiber serves multiple endpoints by using passive fiber optic splitters to divide the fiber bandwidth between the endpoints.

[0007] Optical line terminal (OLT) 14 is located in a central office that interfaces directly or indirectly with core network 15. The interface 16 between OLT 14 and core network 15 may be one or more optical fibers or any other type of communication medium. OLT 14 forms optical signals for transmission downstream to ONTs 11 through feeder optical fibers 17 and also receives optical signals from ONTs 11 through feeder optical fibers 17. Optical splitter 12 is typically a passive device that distributes signals received from OLT 14 to ONTs 11. Similarly, optical splitter 12 receives optical signals from ONTs 11 and provides optical signals to OLT 14 through feeder optical fibers 17. In this manner, a PON includes an OLT with multiple ONTs, which reduces the amount of fiber required compared to a point-to-point architecture.

[0008] As may be observed, an optical signal containing all data for ONT 11 is provided onto feeder fiber 17. As a result, all data provided to each of the ONTs is provided to all ONTs through optical splitter 12. Each ONT selects the portion of the received optical signal intended for that particular ONT and passes the data along to the subscriber, while discarding the remaining data. Typically, data for the ONTs is broadcast onto feeder fiber 17 and provided to each of the ONTs.

[0009] Upstream transmissions from the ONTs 11 through their respective optical fibers 13 are typically transmitted in bursts according to a schedule provided to each ONT by the OLT. In this manner, each of the ONTs 11 transmits upstream optical data at different times. In some embodiments, the upstream and downstream transmissions are transmitted using different wavelengths of light so that they do not interfere with each other. In this manner, the PON may utilize wavelength division multiplexing, using one wavelength for downstream traffic and another wavelength for upstream traffic over a single-mode fiber.

[0010] A schedule from the OLT allocates upstream bandwidth to ONTs. Because the optical distribution network is shared, ONT upstream transmissions are likely to collide if they are transmitted at random times. ONTs are typically located at various distances from the OLT and / or optical splitter, resulting in different transmission delays from each ONT. The OLT measures the delay and sets registers in each ONT to equalize that delay for the OLT and other ONTs associated with it. Once the delay is addressed, the OLT transmits so-called grants to individual ONTs in the form of grant maps. A grant map is an authorization to use a defined time interval for upstream transmission. The grant map is dynamically recalculated periodically, such as for each frame. The grant map allocates bandwidth to all ONTs, ensuring that each ONT receives a timely bandwidth allocation for its service needs. Much data traffic, such as website browsing, tends to be bursty and fluctuates significantly over time. Due to dynamic bandwidth allocation (DBA) among different ONTs, the PON may be oversubscribed for upstream traffic. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] For a better understanding of the present invention and to show how the same may be carried into effect, reference will now be made, by way of example, to the accompanying drawings in which: [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 1 illustrates a network that includes a passive optical network. [Figure 2] FIG. 2 illustrates a passive optical network with downstream data traffic. [Figure 3] FIG. 3 illustrates a passive optical network with upstream data traffic. [Figure 4] FIG. 4 illustrates a remote OLT. [Figure 5] FIG. 5 illustrates an exemplary OLT. [Figure 6] Figure 6 illustrates the processing for a YANG data model using OMCI. [Figure 7] FIG. 7 illustrates the process for a PON network. [Figure 8] Figure 8 illustrates a YANG request and response. [Figure 9] FIG. 9 illustrates a PON network with a remote OLT. [Figure 10] FIG. 10 illustrates a PON network with an OLT. [Figure 11] FIG. 11 illustrates failover for an OLT. [Figure 12] FIG. 12 illustrates a failover switchover. [Figure 13] FIG. 13 illustrates OLT failover. [Figure 14] FIG. 14 illustrates the OLT failover distribution. [Figure 15] FIG. 15 illustrates OLT ranging. [Figure 16] FIG. 16 illustrates OLT reranging. DETAILED DESCRIPTION OF THE INVENTION

[0013] Referring to Figure 2, PON networks are based on a point-to-multipoint downstream transmission arrangement. Data from the OLTs is transmitted to all ONTs interconnected thereto. The data from the OLTs is transmitted in the form of one or more frames, where each frame contains data for one or more of the ONTs. For example, in GPON, a constant of 125 μs frames is used, where each frame contains an allocation map that informs (among other control information) about the slots assigned to the allocation ID. As a result, each frame is divided into one or more time slots designated for a corresponding selected one of the ONTs.

[0014] Referring to Figure 3, a PON network is based on a multipoint-to-point upstream transmission arrangement using a time-division multiple access mechanism. The OLT allocates a time slot (BW map) to each ONT to transmit its upstream transmission, ensuring collision-free transmission. Data from each ONT is transmitted to a corresponding OLT interconnected to the ONT. Data from the ONT is transmitted in the form of a portion of one or more frames, where each frame contains data for one or more of the ONTs. For example, in GPON, a reference frame of 125 μs is used, but this is not an absolute value because an allocation round may span multiple upstream frames. GPON uses a generic encapsulation scheme (GEM), which allows the transport, segmentation, and reassembly of Ethernet frames and legacy traffic (ATM or TDM). As a result, each frame is divided into one or more time slots designated for a corresponding selected one of the ONTs.

[0015] Referring to FIG. 4, in some installations, it is often desirable to locate an optical line terminal at a location remote from the core network, commonly referred to as a remote optical line terminal (OLT). The remote OLT may include one or more feeds from the core network to the remote OLT. The remote OLT may then distribute data to and receive data from multiple ONTs. Each of the ONTs then provides data to and receives data from customer devices. A remote OLT typically has the capacity to provide service to hundreds to thousands of ONTs.

[0016] Referring to Figure 5, an exemplary OLT is illustrated, which may include a local or remote OLT. By way of example, the diag process, dma process, clish process, restapi process, gRPC process, rolt4isr process, and / or rolt4api process are preferably contained locally on the OLT. Also, a dynamic bandwidth allocation process that allocates available bandwidth among each of the ONTs is similarly contained locally on the OLT. Other processes associated with a remote OLT, such as the vomci and / or Yuma servers, may be virtualized and located on cloud-based servers. For example, the VOMCI may (1) receive service configurations from the vOLT management function, (2) translate the received configurations to ITU G.988 OMCI management entities and format them into OMCI messages, (3) encapsulate and transmit the formatted OMCI messages to and from the VOMCI proxy, (4) translate OMCI messages (presenting operational data of the ONT) received from the vOMCI proxy into data understandable by the vOLT management function (e.g., notification acknowledgments, alarms, PM registers), and / or (5) transmit the above ONT operational data to the vOLT management function. See TR-451 vOMCI Specification, June 2022, and ONT management and control interface (OMCI) specification, G.988, November 2017, both of which are incorporated herein by reference in their entireties.

[0017] As an example, gRPC interfaces with multiple vomci agents, which may provide a gRPC server and client layer to provide vomci services to ROLT.

[0018] As an example, the dispatcher provides a messaging pathway between components within the ponapp. A local microservice may register a callback for a message ID that is part of the MSG layer. Any microservice can route to another based on the top two bytes of the message ID, which indicates the destination.

[0019] As an example, IPC provides TCP and UDP sockets for relaying messages to and from applications in MSG lib format, and runs side-by-side with the dispatcher.

[0020] As an example, mgm is a ranging manager that provides the state machine and local for physical layer management of ONTs, including the auto-discovery process, ranging of ONTs, drift management, and LOS handling.

[0021] As an example, shwm is the shelf manager task that handles any device outside of the rolt4api / rolt4isr domain.

[0022] As an example, rolt4isr is a handler for interrupts coming from the PL.

[0023] As an example, rolt4api handles requests from various microservices within ponapp to programmatically interact with or with ROLT.

[0024] As an example, sim provides a simulation service, providing the ability to simulate devices that may not physically exist.

[0025] As an example, spit is a smart card proxy interface task that provides a server for application requests coming into or leaving ponapp. A typical path would be from the outside client through the dispatcher via IPC into spit. SPIT may then relay to other microservices to perform the requested task. Some provisioning may go through the softlib DB and be further relayed as provisioning callouts.

[0026] As an example, mntc is a maintenance state machine, which is preferably an event-driven state machine of the ONT.

[0027] As an example, fdi is a fault detection and isolation task that serves as a hierarchical alarm tree for tracking alarm conditions for different equipment types.

[0028] As an example, stat is a statistics manager that handles polling of on-board statistics and aggregating statistics for other calling functions.

[0029] As an example, iptv provides an IPTV service that includes IGMP snooping / proxy support.

[0030] As an example, dapr is a destination address programmer that handles unknown upstream source mac addresses for N:1 connections. It not only maintains a mac forwarding table in PL, but may also remove stale mac addresses.

[0031] As an example, iotm is an IOT (aka ONT) manager that supports commands to ONTs.

[0032] As an example, dba is dynamic bandwidth allocation.

[0033] As an example, keyx is a key exchange task that handles key exchange for ONTs.

[0034] As an example, softlib is a soft DB library implemented as a memory-based database used to contain the configuration of ROLT.

[0035] As an example, ponid is a library used to associate ITUT serial numbers with ONT IDs and / or channel allocations.

[0036] As an example, debug is the debug library.

[0037] As an example, trans is a transaction library used for transactional and state-based requests to microservices.

[0038] As an example, QBList is a library of various list and vector functions.

[0039] As an example, LOG is an event log.

[0040] As an example, MSG is a message library.

[0041] As an example, QB_OS is an operating system library.

[0042] As an example, QBLIB is a library of local APIs.

[0043] As an example, TIME is a timer library used for time-based callback logic.

[0044] As an example, PLMM is a ploam message library used to encode and decode ploam messages over pon.

[0045] The core network and / or optical line terminals provide management and control functions for the ONTs by using an optical network unit management and control interface (OMCI). The core network 200 and the OLTs 210, through which the core network 200 provides and receives data, transmit and receive data using PON protocols over an optical distribution network (e.g., optical splitter) 220. The OLTs 210 pass data to and receive data from the ONTs 230 through the optical distribution network (ODN) 220. OMCI messages between the ONTs 210 and ONUs 230 for management and control are also provided between the OLTs 210 and ONTs 230 through the ODN 220. The ONTs 230 provide access network line termination, user network interface line termination for subscriber devices, and service multiplexing and demultiplexing for subscriber devices.

[0046] Configuration management provides the ability to identify ONT capabilities and exercise control over the ONT. Management areas for the ONT include the configuration of (1) equipment, (2) passive optical network and reach extender protection, (3) user-network interfaces, (4) gigabit-capable passive optical network encapsulation port network contention termination points; (5) interworking termination points; (6) operations, management, and maintenance flows; (7) physical ports; (8) gigabit-capable passive optical network encapsulation conformance layer profiles; (9) service profiles; (10) traffic descriptors; and (11) asynchronous transfer mode conformance layer profiles. As modeled by OMCI, the ONT detects and reports equipment, software, and interface faults and declares corresponding alarms. The ONT may be considered a managed entity through the exchange of information between the OLT and the ONT based on OMCI messages to the optical access network.

[0047] Each of the capabilities and management functions of an ONT is described in a more or less concise manner by various standards, which are typically achieved by consensus among a diverse set of entities, each of which tends to have a different perspective on the meaning of the standard's descriptions. Thus, each ONT, and especially ONTs developed by different manufacturers, may have variations based on that particular manufacturer's interpretation of the various standards. This tends to be particularly true for control and management functions.

[0048] The G.988 standard describes management entities in a protocol-independent Management Information Base (MIB) that models the exchange of information between OLTs and ONTs in PON-based access networks that are the subject of standards such as G.988. See G.988: ONU management and control interface (OMCI) specification, (November 17); G.988 (2017) Amendment 1 (November 18); G.988 (2017) Amendment 2 (August 19); G.988 (2017) Amendment 3 (March 2); and G.988 (2017) Amendment 4 (September 21), each of which is incorporated herein by reference in its entirety. G.988 also addresses ONT management and control channel (OMCC) setup, protocols, and message formats. Additionally, considerations of various manufacturer interpretations of the G.988 standard often are not sufficient for complete interoperability between different OLT and ONT manufacturers. Due to manufacturer decisions on implementation, various ONTs exist that simply do not comply with the various standards.

[0049] Referring to Figure 6, one technique for providing OMCI messages to ONTs is to create a virtual OMCI set of microservices tailored to the functionality for each ONT model from each vendor for a server in the core network (i.e., any server in the network). The management data maintained by the system is typically defined in terms of a YANG data model, including modules and submodules that define configuration and status data, notifications, and therefore remote procedure calls. A YANG module defines a data model through its data and through the hierarchical organization of and constraints on that data. Each module is uniquely identified by a namespace URI. A module defines a single data model. However, a module can reference definitions in other modules and submodules by importing external modules using the import statement or by including one or more submodules using the include statement. Additionally, a module can extend another data model by using the augment statement to define the placement of a new node in the data model hierarchy and the when statement to define the conditions under which the new node becomes valid. Modules use feature statements to identify parts of the module that are conditional and deviation statements to identify places where the device's implementation may deviate from the original definition. In this way, modules can have a large and complex set of conditions to suit different environments. The core network provides YANG requests to the OLT, which then translates the YANG requests and responses and notifications to and from the vOLTMF (vOLT Management Function) into OMCI messages, and the OLT sends and receives OMCI message requests and responses and notifications to and from the ONT.

[0050] Referring to Figure 7, a high-level design of the vOLT Management Function (vOLTMF) is illustrated, which may be used to manage ONTs through vOMCI messages. Communication occurs between the vOLTMF, vOMCI proxy, and vOMCI function based on the creation and deletion of ONTs, receiving ONT state change notifications, and sending requests to ONTs. The vOLTMF manages ONTs through an ONT adapter, which may be deployed as a broadband access abstraction, and its association is based on the model, type, vendor, and version stated during the ONT's creation. The ONT adapter may use a library of YANG modules for the ONT referenced by the vOLTMF to handle ONT requests, responses, and notifications from external management systems.

[0051] The vOLTMF performs actions upon receiving notifications and requests from either the OLT device or other components within the broadband access abstraction core. For example, an onu state change notification sent by the OLT device on its northbound interface (NBI) is received by the broadband access abstraction core. The broadband access abstraction core propagates the notification to the vOLTMF and the broadband access abstraction NBI so that it can be handled by the access SDN M&C.

[0052] Upon receiving the notification, vOLTMF processes the notification, checks whether the pre-configured ONU device exists, and authenticates the ONU. vOLTMF transforms the notification into Google Protobuf (GPB) format and propagates a set-onu-communication action towards the vOMCI function and vOMCI proxy via the Kafka bus.

[0053] All YANG requests are sent to the vOMCI function and vOMCI proxy via the Kafka bus in GPB format. Once the vOMCI function / proxy processes the request, the vOMCI function sends a notification / request response back to vOLTMF in GPB format via the Kafka bus, and the response is received through KafkaNotificationCallback#onNotification().

[0054] Upon receiving the response, the vOLTMF is responsible for processing the response and taking action accordingly.

[0055] There can be multiple interactions between vOLTMF and vOMCI functions, including parallel configuration requests / commands to either the same or different ONUs. These interactions are parallel and asynchronous, which allows vOLTMF to have separate task queues and thread pools to handle request / response interactions so that requests are not idle / blocked while waiting for a response. Below is a list of vOLTMF thread pools spawned as new runnable tasks: processNotificationRequestPool, kafkaCommunicationPool, kafkaPollingPool, processNotificationResponsePool, and processRequestResponsePool. processNotificationRequestPool is used to process mediated device event listener callbacks and device notification requests. kafkaCommunicationPool is used to process individual GET / COPY-CONFIG / EDIT-CONFIG requests inside the MediatedDeviceNetconfSession spawned by preocessRequestResponsePool. The kafkaPollingPool is used to fine-tune the KafkaConsumer implementation and poll for responses from the vOMCI function / vOMCI proxy. The processRequestResponsePool is used to handle notification responses from the vOMCI function / vOMCI proxy. The processRequestResponsePool is used to handle GET / COPY-CONFIG / EDIT-CONFIG requests and responses from the vOMCI function / vOMCI proxy. In general, processes can be thought of as a type of protocol adapter for those running on the ONT, which also works with the OLT in a PON environment.As may be observed, the manner in which the processing is performed is relatively complex, involving Google Protobufs, remote procedure calls, and other complexities that require a significant amount of computing resources to process all the microservices burdened on the OLT.

[0056] Referring to Figure 8, generally, a server constructs or otherwise selects a YANG request for an ONT. The server then provides the YANG request to the OLT, which converts the YANG request into an OMCI message and sends the OMCI message to the ONT. The OLT receives the OMCI messages from the ONT and converts them into a YANG response that is provided to the server.

[0057] Referring to FIG. 9, in many access networks, there are multiple remote OLTs interconnected to a core network, each of which provides service to a corresponding set of ONTs.

[0058] Referring to FIG. 10, in many access networks, there are multiple OLTs co-located and interconnected to a core network, each of which provides service to a corresponding set of ONTs.

[0059] The OLT may include a data plane that controls how data is processed, including dynamic bandwidth allocation located on each OLT. The OLT may include a control plane that provides control over how the data plane processes data, such as information in routing tables that define what the data plane does with incoming data. The OLT may include a management plane that configures, monitors, and provides management, monitoring, and configuration services to the control and data planes. The control plane, or portions thereof, for a particular OLT may be virtualized and executed by the core network, with the remaining control plane executed by the OLT. For example, a vOLT and / or vOMCI may be virtualized and executed by the corresponding core network. The management plane, or portions thereof, for a particular OLT may be virtualized and executed by the core network, with the remaining management plane executed by the OLT. The data plane, or portions thereof, for a particular OLT may be virtualized and executed by the core network, with the remaining data plane executed by the OLT. Preferably, most, if not all, of the data plane is executed by the OLT. Preferably, most, if not all, of the control plane is executed by the core network. Preferably, most, if not all, of the management plane is executed by the core network.

[0060] Referring to FIG. 11 , in many environments, it is desirable to include failover protection for a primary OLT in the event that the primary OLT stops providing reliable service to the ONT by including a redundant OLT, which may be included in the same chassis or in a separate chassis. When the primary OLT stops providing service, including reliable service, to the ONT, the service provided by the primary OLT may be stopped, and the redundant OLT may be used to provide service to the ONT. The primary OLT may include a primary laser that provides an optical signal to the ONT onto an optical fiber and a primary optical sensor that senses the optical signal from the ONT from the optical fiber. The redundant OLT may include a redundant laser that provides an optical signal to the ONT onto the optical fiber and a redundant optical sensor that senses the optical signal from the ONT from the optical fiber. When the primary laser is providing a signal to the ONT, the redundant laser does not provide a signal to the ONT. Also, when the primary optical sensor senses the optical signal from the ONT, the redundant optical sensor does not sense the signal from the ONT. Optical signals from the primary and redundant lasers to the ONT may be provided by an optical combiner, and optical signals from the ONT to the primary and redundant optical sensors may be provided by an optical splitter. A controller may be used to control which primary and redundant components are used, such as based on connectivity with the ONT. Preferably, a primary laser transmitter / primary optical sensor or a redundant laser transmitter / redundant optical sensor is used, although one of each may be used depending on the configuration. Preferably, the controller is included with the ONT, but may also be included as part of the core network. As may be observed, the primary and redundant optical paths share the same optical fiber from the optical combiner and / or optical splitter to the ONT, and they each have separate optical fibers from their respective lasers and optical sensors to the optical splitter / combiner. The input from the core network to the OLT is preferably an Ethernet packet-based interconnect.

[0061] If the primary OLT stops operating properly or the respective ONTs do not otherwise communicate properly with the primary OLT, the network operator may selectively disable the primary OLT and enable the redundant OLT to provide service to the ONTs, such as by signaling the controller. Preferably, the controller and / or core network determines that the primary OLT stops operating properly or the respective ONTs do not otherwise communicate properly with the primary OLT, selectively disable the primary OLT, and enable the redundant OLT to provide service to the ONTs. The controller and / or core network may also determine that the redundant OLT stops operating properly or the respective ONTs do not otherwise communicate properly with the redundant OLT, selectively disable the redundant OLT, and enable the primary OLT to provide service to the ONTs.

[0062] Referring to FIG. 12 , the origin of a fault can result from a number of different sources, such as the primary component, the redundant component, the optical fiber from the primary component to the optical splitter / combiner, the optical fiber from the redundant component to the optical splitter / combiner, one of the primary components (e.g., the primary laser transmitter or the primary optical sensor), one of the redundant components (e.g., the redundant laser transmitter or the redundant optical sensor), a portion of the optical splitter / combiner, a control component for the primary laser transmitter or the primary optical sensor, and / or a control component for the redundant laser transmitter or the redundant optical sensor. To provide a more effective data connection with the device, the controller may monitor whether the primary component is effectively communicating with some or all of the ONTs provisioned on the passive optical network for the corresponding optical line transmitter 1200. If the data connection to the ONT with the primary component is not effective, such as if some or all of the provisioned ONTs are unavailable 1210, the controller switches communication to the ONT using the redundant component while switching off the primary component 1220. The controller may monitor 1230 whether the redundant component is effectively communicating with some or all of the ONTs provisioned on the passive optical network for the corresponding optical line transmitter. The controller preferably monitors 1230 whether the redundant component is effectively communicating with some or all of the ONTs provisioned for a period of time, preferably greater than 5 seconds, preferably greater than 10 seconds, more preferably greater than 20 seconds, and more preferably greater than 30 seconds. If, after the elapsed time period, the controller determines that connectivity with the provisioned ONTs using the redundant component is ineffective, such as because some or all of the provisioned ONTs are unavailable 1240, the controller switches 1250 communication to the ONTs using the primary component while switching off the redundant component.The controller preferably monitors whether the primary component is effectively communicating with some or all of the provisioned ONTs for a period of time, preferably greater than 5 seconds, preferably greater than 10 seconds, more preferably greater than 20 seconds, and more preferably greater than 30 seconds 1200. As a result, the controller manages switching between the primary and redundant components with a time delay between switching until such time as effective communication with the corresponding ONT is established.

[0063] The controller may include, for example, a field programmable gate array (FPGA), which includes blocks of gates that can be configured to implement logic. In comparison, a microprocessor is a central processing unit (CPU) that executes a program containing a specific set of instructions. A microprocessor has a fixed set of instructions used for appropriate programming, commonly referred to as programming code. Each of these instructions has its own corresponding block hardwired to the microprocessor. In comparison, an FPGA does not have such hardwired logic blocks. FPGAs are often laid out like nets, with each junction containing a switch that can be connected or disconnected. This set of interconnections determines how the logic of each block is determined. Programming an FPGA typically involves a hardware description language, commonly referred to as programming logic. In some cases, an FPGA and a microprocessor are combined in a single package, providing additional flexibility. The microprocessor typically performs most of the generalized processing while handing off more specific tasks to the FPGA gate array. For OLTs, the combination of a microprocessor and an FPGA gate array in a single package (i.e., chip) based processing system provides programming flexibility along with specialized logic processing that is particularly well suited to supporting reduced power usage constraints.

[0064] The controller, which includes a processing system (e.g., FPGA and microprocessor-based chips), preferably includes most of the data plane processing, including dynamic bandwidth allocation. The controller may also include portions of the control plane and / or management plane, while most of the control plane and / or management plane are virtualized and provided by the core network.

[0065] As an example, most of the processing side of the processing system is preferably virtualized to a computing device external to the OLT (e.g., the core network). Preferably, the virtualized processing side of the processing system is primarily related to the control pane, including most of the vOMCI and the vOLT. While most of the programming logic of the processing system is preferably data plane processing for the OLT, a portion of it may be virtualized to a computing device (e.g., the core network). Preferably, the programming logic of the processing system is primarily related to the data plane. In particular, it is desirable to include a dynamic bandwidth allocation process that is contained within the processing system and is not virtualized.

[0066] Referring to FIG. 13 , in one embodiment, an access network for a passive optical network typically includes a set of optical line terminals (OLTs), each of which includes local processing capabilities such as an FPGA, a microprocessor, and / or a processing system. Additionally, a portion of the processing required for each of the OLTs (e.g., the data plane, the control plane, and / or the management plane) may be provided as a virtualized service on the core network. When a primary OLT providing services to each ONT is replaced by a redundant OLT providing services to each ONT, the virtualized services provided by the core network may be provided to the redundant OLT. In this manner, the redundant OLT may reuse existing data from the primary OLT, reducing the computational burden on the redundant OLT.

[0067] Referring to FIG. 14 , in another embodiment, an access network for a passive optical network typically includes a set of OLTs, each of which includes local processing capabilities such as an FPGA, a microprocessor, and / or a processing system. Also, a portion of the processing required for each OLT (e.g., the data plane, the control plane, and / or the management plane) may be provided as a virtualized service on the core network. If a primary OLT providing services to each ONT is replaced by a redundant OLT providing services to each ONT, the services provided to the replaced primary OLT may be provided by the redundant OLT. Also, if a primary OLT providing services to each ONT is replaced by a redundant OLT providing services to each ONT, the services provided to the replaced primary OLT may be provided by another OLT in the network, since processing capacity is available. In this way, a collection of OLTs may collectively provide redundancy and processing capacity for other OLTs, using at least a portion of the services for a particular OLT provided by another OLT.

[0068] There are different distances between the OLT and the ONT due to their physical locations relative to each other. The different distances result in different times required for an optical signal from the OLT to reach the ONT and for an optical signal from the ONT to reach the OLT. Ranging may be used to reduce data collisions for different ONTs. For example, ranging techniques may enable timing adjustment between the OLT and each of the ONTs, where each of the ONTs may have a different temporal range offset. The OLT initiates ranging by sending a specific grant and opens a window to receive a response from the ONT. After receiving the grant, the ONT sends a ranging call back to the OLT. The OLT assigns an equalization delay to a specific ONT based on the elapsed time between sending the ranging grant and receiving the response.

[0069] Referring to FIG. 15, ranging is necessary to avoid upstream transmission collisions among numerous ONTs. Because ONTs are located at different distances from the OLT, data should arrive at the OLT within the correct time slot to reduce errors. The ranging mechanism creates a required transmission delay. Ranging is initiated by sending a ranging request message from the OLT to a specific ONT. Based on the response and the RTD (round trip delay), an equalized delay is calculated. This information is then forwarded toward the ONT, where the ONT is located at the same virtual distance from the OLT as other ONTs. Ranging is performed for all corresponding ONTs. Due to ranging, data should not have any transmission conflicts. However, until all ONTs have been ranged, the time required to perform ranging for all corresponding ONTs is quite long because ONTs tend to transmit simultaneously, requesting that a selected ONT withdraw while other ONTs transmit.

[0070] For a primary OLT, there is a distribution of ranging times that is determined and provided to corresponding ONTs. Periodically, due to temperature and other changes over time, the OLT re-ranges selected ONTs and selectively updates the ONTs with updated ranging information. If the primary OLT fails or otherwise does not provide effective communication with the ONTs, the system switches to a redundant OLT. A redundant OLT is not suitable for direct use of the ranges used by the primary OLT because even for OLTs with seemingly the same effective optical distance, the ranging information tends to be sufficiently different that it is not suitable for direct use by the redundant OLT. Furthermore, even relative differences in the ranging information of the primary OLT tend to be sufficiently different for at least some of the OLTs that it is not suitable for direct use by the redundant OLT.

[0071] Referring to FIG. 16, the primary OLT ranging is not sufficiently accurate for redundant OLT ranging, but it is sufficient to provide useful information that may be used for ranging by the redundant OLT. As an example, the redundant OLT may select a subset of ONTs for re-ranging, such as those that are likely to be consistent with each other. As an example, the redundant OLT may select a set of ONTs with sufficiently different ranging values ​​so that a sufficient time interval exists between ranging. The remaining ONTs are likely to have a generalized offset from the ranged set of ONTs by the redundant OLT based on the offset of the primary OLT. For example, if ONT 15 has a range value of 1000 with respect to the primary OLT, it may have a range value of 1004 with respect to the redundant OLT. For example, if ONT 16 is determined to have a range value of 1003 (+3 with respect to ONT 15), the redundant OLT may initially estimate that ONT 16 is likely to have a range of approximately 1007 (1004 + 3 with respect to ONT 15). The redundant OLT may estimate ranging for the remaining ONTs based on an initial ranging set of one or more ONTs. Based on the estimate of the remaining ONTs (or set), the redundant OLT may select a set of ONTs, or the remaining ONTs, to range with, along with the interval between sending requests for ranging to the selected ONTs, thereby substantially reducing the likelihood that the ONTs will interfere with each other during ranging. The redundant OLT may continue this process until all ONTs have been ranged. The selective use of ranging information from the primary OLT for the redundant OLT, along with the timing of the ranging requests, substantially reduces the time required to perform ranging.

[0072] Other techniques for using ranging information from the primary OLT to inform ranging information for the redundant OLT to reduce the time to perform such ranging may be used as well.

[0073] Furthermore, each functional block or various features in each of the foregoing embodiments may be implemented or performed by a circuit, typically an integrated circuit or multiple integrated circuits. A circuit designed to perform the functions described herein may include a general-purpose processor, a digital signal processor (DSP), an application-specific or general-purpose integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, or discrete hardware components, or a combination thereof. A general-purpose processor may be a microprocessor, or alternatively, the processor may be a conventional processor, controller, microcontroller, or state machine. The general-purpose processor or each circuit described above may be composed of digital or analog circuits. Furthermore, if advances in semiconductor technology result in the emergence of a technology for fabricating integrated circuits that replace current integrated circuits, integrated circuits based on this technology may also be used.

[0074] It is to be understood that the present invention is not limited to the particular embodiments described, and that modifications may be made therein without departing from the scope of the invention as defined in the appended claims, as interpreted in accordance with general principles, including the doctrine of equivalents or any other principle that extends the scope of binding claims beyond their literal scope. Unless the context dictates otherwise, a reference in a claim to the number of instances of an element, whether to one instance or to more than one instance, requires at least the recited number of instances of the element, but is not intended to exclude from the scope of the claim structures or methods having more instances of that element than are recited. When used in the claims, the word "comprise" or derivatives thereof are used in a non-exclusive sense that is not intended to exclude the presence of other elements or steps in the claimed structure or method.

Claims

1. 1. An access network for a passive optical network, comprising: (a) a first optical line terminal including a northbound interface capable of receiving and transmitting data from and to a server, respectively; (b) said first optical line terminal including ports each capable of receiving and transmitting optical data from and to a set of optical network terminals over optical fibers; (c) a second optical line terminal including a northbound interface capable of receiving data from and transmitting data to said server, respectively; (d) said second optical line terminal including ports each capable of receiving and transmitting optical data from and to said set of optical network terminals over said optical fiber; (e) a controller that determines whether one of the first optical line terminal and the second optical line terminal currently attempting to transmit and receive the optical data is not in effective communication with one of the optical network terminals, and in response, switches to the other of the first optical line terminal and the second optical line terminal to attempt to transmit and receive the optical data using one of the optical network terminals, the controller determining the switch based, at least in part, on monitoring the effective communication during a temporal period of at least 5 seconds.

2. The access network of claim 1 , further comprising: the temporal period being at least 10 seconds.

3. The access network of claim 1 , further comprising: the temporal period being at least 20 seconds.

4. The access network of claim 1 , further comprising: the temporal period being at least 30 seconds.

5. 2. The access network of claim 1, further comprising: the controller determining that the other of the first optical line terminal and the second optical line terminal currently attempting to transmit and receive the optical data is not in effective communication with one of the optical network terminals, and in response, switching to the other of the first optical line terminal and the second optical line terminal to attempt to transmit and receive the optical data using one of the optical network terminals, and the controller determining the switching based, at least in part, on monitoring the effective communication during a temporal period of at least 5 seconds.

6. 1. An access network for a passive optical network, comprising: (a) a first optical line terminal including a northbound interface capable of receiving and transmitting data from and to a server, respectively; (b) the first optical line terminal including ports capable of respectively receiving and transmitting optical data from and to the first set of optical network terminals over a first optical fiber; (c) a second optical line terminal including a northbound interface capable of receiving data from said server and transmitting data to said server, respectively; (d) said second optical line terminal including ports capable of respectively receiving and transmitting optical data from and to said first set of optical network terminals over said first optical fiber; (e) a third optical line terminal including a northbound interface capable of receiving data from and transmitting data to said server, respectively; (f) a third optical line terminal including ports capable of receiving and transmitting optical data from and to a second set of optical network terminals, respectively, over a second optical fiber, wherein none of the first set of optical network terminals is included in the second set of optical network terminals; (g) a controller that determines whether one of the first optical line terminal and the second optical line terminal currently attempting to transmit and receive the optical data is not effectively communicating with one of the optical network terminals, and in response, switches to the other of the first optical line terminal and the second optical line terminal and attempts to transmit and receive the optical data using one of the optical network terminals, wherein the access network reallocates at least a portion of at least one of the control plane processes and management plane processes provided by the first optical line terminal to the third optical line terminal.

7. 7. The access network of claim 6, wherein the access network relocates at least a portion of both the control and management plane processes provided by the first optical line terminal to the third optical line terminal.

8. 1. An access network for a passive optical network, comprising: (a) a first optical line terminal including a northbound interface capable of receiving and transmitting data from and to a server, respectively; (b) the first optical line terminal including ports capable of respectively receiving and transmitting optical data from and to the first set of optical network terminals over a first optical fiber; (c) a second optical line terminal including a northbound interface capable of receiving data from and transmitting data to said server, respectively; (d) said second optical line terminal including ports capable of respectively receiving and transmitting optical data from and to said first set of optical network terminals over said first optical fiber; (e) a controller that determines whether one of the first optical line terminal and the second optical line terminal currently attempting to transmit and receive the optical data is not effectively communicating with one of the optical network terminals, and in response, switches to the other of the first optical line terminal and the second optical line terminal to attempt to transmit and receive the optical data using one of the optical network terminals, wherein the server includes a virtualized control plane service for the first optical line terminal, and the server is switched to provide a virtualized control plane service for the second optical line terminal.

9. 9. The access network of claim 8, further comprising the server including a virtualized management plane service for the first optical line terminal, the server being switched to provide a virtualized control plane service for the second optical line terminal.

10. 1. An access network for a passive optical network, comprising: (a) a first optical line terminal including a northbound interface capable of receiving and transmitting data from and to a server, respectively; (b) said first optical line terminal including ports each capable of receiving and transmitting optical data from and to a set of optical network terminals over optical fibers; (c) a second optical line terminal including a northbound interface capable of receiving data from and transmitting data to said server, respectively; (d) said second optical line terminal including ports each capable of receiving and transmitting optical data from and to said set of optical network terminals over said optical fiber; (e) a controller that determines whether one of the first optical line terminal and the second optical line terminal currently attempting to transmit and receive the optical data is not effectively communicating with one of the optical network terminals, and in response, switches to the other of the first optical line terminal and the second optical line terminal to attempt to transmit and receive the optical data using one of the optical network terminals, wherein the second optical line terminal re-ranges the set of optical network terminals based on ranging data from the first optical line terminal.

11. 11. The access network of claim 10, wherein the second optical line terminal re-ranges at least one of the optical network terminals of the set of optical network terminals and uses a change in the ranging of the one of the optical network terminals between the first optical network terminal and the second optical network terminal to modify the timing of the re-ranging of another one of the set of optical network terminals.

Images