US20130188946A1 - System for redundancy in ethernet passive optical networks (epons) - Google Patents
System for redundancy in ethernet passive optical networks (epons) Download PDFInfo
- Publication number
- US20130188946A1 US20130188946A1 US13/794,834 US201313794834A US2013188946A1 US 20130188946 A1 US20130188946 A1 US 20130188946A1 US 201313794834 A US201313794834 A US 201313794834A US 2013188946 A1 US2013188946 A1 US 2013188946A1
- Authority
- US
- United States
- Prior art keywords
- olt
- onu
- olts
- switchover
- redundancy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0062—Network aspects
- H04Q11/0067—Provisions for optical access or distribution networks, e.g. Gigabit Ethernet Passive Optical Network (GE-PON), ATM-based Passive Optical Network (A-PON), PON-Ring
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/03—Arrangements for fault recovery
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/03—Arrangements for fault recovery
- H04B10/032—Arrangements for fault recovery using working and protection systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04J—MULTIPLEX COMMUNICATION
- H04J14/00—Optical multiplex systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0062—Network aspects
- H04Q2011/0079—Operation or maintenance aspects
- H04Q2011/0081—Fault tolerance; Redundancy; Recovery; Reconfigurability
Definitions
- the present embodiment generally relates to the field of communications networks, and in particular, concerns a system for redundancy in Ethernet passive optical networks.
- FIG. 1 a diagram of an EPON network
- an OLT optical line transmission equipment of the network provider
- 100 , 101 communicates over a fiber optic network
- ONUs optical networking units associated with a user
- a passive splitter 106 facilitates either OLT-0 100 or OLT-1 101 communicating with the ONUs 108 A, 108 B, 108 N.
- a host 124 provides configuration information, command, control, and monitoring of the OLTs ( 100 , 101 ).
- a variety of architectures can be used to provide redundancy of the fiber optic network and the EPON communications standard includes features to facilitate recovery from a network failure. Redundancy in an EPON system facilitates continuity of service including switchover from a first OLT (sometimes called the “main”) to a second OLT (sometimes called the “standby”) and/or from a first network to a second network. This switchover can include alternative OLTs and fiber optic cables.
- An ONU will deregister under certain conditions, including:
- a long time normally greater than 300 msec, has passed without the ONU receiving any downstream GATE packets.
- a new OLT not having the current security key for an ONU When an OLT sends data using a security key that an ONU is no longer using, the ONU will not recognize the data. Because the ONU does not recognize the data, the ONU will be in a state of not receiving data, and after a given time has elapsed without the ONU receiving data, the ONU will deregister. In a case where the control packets are not encrypted, the ONU will still not be able to recognize the data, and will be in a state of not receiving data. In the context of this document, data is used as a general reference to all information communicated, unless otherwise specified. In a case where the data that the system secures is only the data packets, but the control packets are sent unencrypted, the ONU will similarly be in a state of not receiving data.
- OLTs typically co-locate the redundant OLTs in order to share common resources, synchronize OLTs, or minimize differences in the length of fiber optic cables.
- a popular implementation of OLTs is as a printed circuit board, with multiple OLTs deployed in a single chassis at a network provider's facility and a single host controlling the OLTs.
- the timestamp counters are not synchronized.
- the timestamp counters advance every 16 nsec (1 TQ), and synchronization requires at least 2 wires (1 wire for clock and 1 wire for reset) connected between the OLTs. This case is a difficult implementation with strict physical limitations.
- RTT round trip delay
- co-locating OLTs and hosts exposes the system to failure due to lack of location redundancy.
- location redundancy refers to deploying portions of the system at physically diverse locations.
- a main and a standby OLT are both located in the same room.
- a natural disaster, such as flooding, an accident, such as a fire, or power shutdown for routine maintenance affect both the main and standby OLTs resulting in a failure of the EPON.
- a system for redundancy in Ethernet passive optical networks includes: at least one optical networking unit (ONU), wherein when the ONU is in a normal state the ONU monitors input communications, and wherein when the input communications are quiet for a predetermined minimum length of time the ONU transitions to a lenient state in which: the ONU accepts old and new security keys; upon receiving a packet: the ONU updates an ONU timestamp based on said packet's timestamp; and the ONU transitions to said normal state of operation; and wherein while the ONU is in the lenient state, if a packet is not received for a predetermined given length of time the ONU transitions to a deregistered state.
- ONU optical networking unit
- the system further includes at least two optical line transmission equipment (OLT) wherein a first OLT is deployed at a first location and a second OLT is deployed at a second location: and a passive optical splitter for exchanging packets between the OLTs and the at least one ONU.
- OLT optical line transmission equipment
- the OLTs are configured to transmit a switchover signal and the OLTs are configured to receive a switchover signal from other OLTs.
- the OLT executes a switchover process.
- the OLT is configured such that after the switchover process IGMP and MAC address table information is determined.
- system is configured to switch from using the first OLT to using the second OLT in less than 50 msec.
- system is configured to recalculate the round trip time (RTT) after OLT switchover while the at least one ONU is still registered.
- RTT round trip time
- FIG. 1 a diagram of an EPON network
- FIG. 2 a diagram of a redundant EPON network
- FIG. 3 a timing flow diagram of the OLT switchover process
- FIG. 4 a flow diagram of ONU lenient mode operation
- FIG. 5 a diagram of an example of an N:1 redundant EPON network.
- the present embodiment is a system for redundancy in Ethernet passive optical networks (EPONs).
- the system facilitates fast recovery from failure, specifically with a recovery time of less than 50 msec.
- the system also facilitates path redundancy of the fiber optic network.
- the system also facilitates implementation with location redundancy of the OLTs, which facilitates greater path redundancy of the fiber optic network.
- Security information for the standby OLT can be updated asynchronously from the main OLT. Note that this system can be used to provide N:1 redundancy. For clarity, the description first describes an implementation of 2:1 redundancy and then describes how this implementation can be extended to provide an N:1 implementation.
- FIG. 2 a diagram of a redundant EPON network
- a first optical line transmission equipment (OLT) 200 A is deployed at a first location and a second OLT 200 B is deployed at a second location.
- OLT optical line transmission equipment
- Deploying each OLT at different physical locations facilitates location redundancy.
- the specific application and desired level of location redundancy will provide requirements for determining what locations will be sufficient to satisfy the location redundancy requirements.
- the OLTs are deployed in the same room, or even in the same chassis, with each OLT having independent power supplies and network connections.
- each OLT is deployed in a different building.
- the location redundancy configuration of OLTs will be limited by the operating parameters of the EPON specification and deployment realities.
- the OLTs are configured to send and receive packets over first communication channels 104 A and 104 B, respectively, to passive optical splitter 106 .
- Second communication channels 107 A, 107 B, and 107 N operationally connect the passive optical splitter 106 to one or more associated optical networking units (ONUs) 214 A, 21413 , and 214 N, respectively.
- ONUs optical networking units
- an active communication channel is a communication channel from which packets are received and to which packets can be transmitted between given devices on the communications channel.
- a quiet communications channel is a communications channel which is not communicating packets between given devices on the communications channel.
- a quiet communications channel can physically have signals being carried on the communications channel and reaching the physical input of a device on the network, but if the signal is not intended for the receiving device, or if the receiving device is not able to process the signal, the communications channel is considered quiet for that device.
- a non-limiting example of a quiet channel is the case where packets are being transmitted from an OLT to an ONU, but the ONU does not have the correct security key to decrypt the received packets. In this case, the ONU is not receiving usable data and the input communications channel is considered quiet.
- the ONUs are configured to monitor an associated second communication channel.
- a conventional ONU transitions from a normal state to a deregistered state, requiring approximately 1000 msec to reregister during which application data is not transmitted or received.
- the ONUs are configured with an innovative “lenient state”.
- the ONU transitions from a normal state to a lenient state.
- the ONU starts accepting both the old (the previously used) security key and the current security key for decrypting received packets, as described below.
- the old security key is used for encrypting and transmitting packets.
- This lenient state allows the ONU to receive a packet, the ONU's timestamp is updated based on the packet's timestamp, and the ONU's security counter is updated based on the packet's security counter.
- the ONU transitions to a normal state of operation.
- the ONU transitions to a deregistered state.
- Using a lenient state reduces the switchover time from an average of 1000 msec to a worst case of 21 msec, as will be described below.
- Hosts 224 A, 224 B are connected to OLT-1 200 A, and OLT-2 200 B by communication channel 234 .
- Network switch 220 is connected to OLT-1 200 A and OLT-2 200 B by communication channels 230 A, and 230 B, respectively.
- the network switch 220 provides connectivity via communications channel 232 to the core network 222 .
- Communications channel 236 between OLT-1 200 A and OLT-2 200 B is for general-purpose input/output pins (GPIO) connection and in the context of this document is referred to as the GPIO connection.
- GPIO output pins from OLT-1 200 A connect to GPIO input pins on OLT-2 200 B.
- GPIO output pins from OLT-2 200 B connect to GPIO input pins on OLT-1 200 A.
- a cable can be used to connect directly to the OLTs' GPIO.
- the GPIO communication channel can be remoted as appropriate to the application. Techniques for remoting communicating channels are known in the art, and based on this description one familiar with the art will be able to implement an appropriate technique.
- the GPIO connection is configured to transmit a switchover signal from the main OLT-1 200 A to the standby OLT-2 200 B. When OLT-2 200 B receives a switchover signal, the switchover process begins for OLT-2 200 B.
- One preferred implementation for transmitting a switchover signal is by lowering the main OLT's GPIO to ‘0’. This can be done by setting the GPIO to input mode where the output gets a ‘high Z’ value and is pulled down by the pull-down resistor.
- the standby OLT receives a switchover signal. This allows the standby OLT to recognize a power failure of the main OLT. Note that as described elsewhere, after the switchover process is complete OLT-2 200 E has become the main OLT and. OLT-1 200 A has become the standby OLT in preparation for future switchovers.
- Fiber protection switching also known as OLT switchover
- OLT switchover can be initiated by a variety of circumstances including:
- the standby and main OLTs can be deployed in a variety of configurations, including, but not limited to, each OLT located in a different physical location, the OLT's co-located but deployed in separate chassis, or OLT printed circuit boards (PCBs) deployed in the same chassis. In a case where the OLTs are implemented on separate printed circuit boards in the same chassis, there is no need to synchronize clock and reset signals between the two boards.
- Each OLT has an associated fiber for communications with the EPON.
- the choice of initially which OLT is the main OLT and which OLT is the standby OLT is dependent on the specifics of the application and deployment requirements. Once in use, either unit can function as the main OLT and the other unit will be the standby OLT. After a switchover is completed, the former main OLT becomes the standby OLT and the former standby OLT operates as the main OLT. It is not necessary to maintain a specific OLT permanently in a specific role as a main or standby OLT.
- OLT-1 200 A we arbitrarily choose OLT-1 200 A, as the initially main OLT and OLT-2 200 B as the initially standby OLT to assist in describing the OLT switchover.
- the standby OLT-2 200 B is passive, not initiating transactions on communications channel 104 B to the EPON or to the network switch via communications channel 230 B.
- one of the hosts 224 A is considered the main host and provides configuration information, command, control, and monitoring of the OLTs.
- Information maintained by the host includes which ONUs are registered and dynamic bandwidth allocation (DBA) order of the ONUs, address table, IGMP proxy table, ONU networking settings, security status (enabled/keys) of each ONU, and static OLT configurations such as classification rules.
- the main host 224 A actively updates and is updated by the main OLT-1 200 A.
- the main host 224 A updates the standby OLT-2 200 B and can update the standby host 224 B.
- column 300 is for actions of the switch, column 310 the host, column 320 the initial standby OLT, and column 340 the initial main OLT.
- the flow down the page is relative time 302 . Note that lengths of time described in this document are preferable, but timeouts and waiting periods are configurable, and the system can be configure with specific lengths of time appropriate to the application.
- the switchover process can be initiated in a variety of ways, including, but not limited to: the host 310 issuing a start switchover command 312 to the main OLT 340 , the main OLT detecting a problem 342 with the fiber communications channel, the main OLT not detecting any upstream traffic, or by detecting a problem with the fiber optic network (for example via fiber monitoring), in which cases the main OLT 340 stops transmitting 344 , by turning off the transmit laser.
- the main OLT 340 triggers 348 the GPIO.
- the switchover process can also be initiated when the main OLT 340 has a power failure 346 , in which case due to loss of power the GPIO is triggered 348 (and the laser turns off due to the power failure).
- the GPIO is triggered according to the pulldown resistor technique described above.
- the switchover command is given to the main OLT to take the main OLT out of use and perform maintenance on the main OLT. In another case, if the main OLT is malfunctioning, the OLT still has power and the GPIO will not be pulled down.
- a malfunctioning OLT is where there is a problem with a processor on the OLT (commonly known as a “stuck” or “hung” CPU).
- a malfunctioning OLT is where there is a problem with the OLT PON receive line, resulting in a quiet communications channel.
- a watchdog process on the host monitors the main OLT, and if the main OLT does not respond to the watchdog process then the host initiates a start switchover command to the standby OLT to start the switchover process.
- the main OLT notifies 350 the host 310 that the main OLT will no longer function as the main OLT.
- the OLT 340 waits 352 a timeout period to allow the original standby OLT 320 to start functioning as the new main OLT and then the OLT 340 becomes the new standby OLT 354 .
- the timeout period should preferably be at least twice the dynamic bandwidth allocation (DBA) cycle to assure that none of the ONUs has to transmit any more upstream packets.
- the DBA cycle is 0.5 msec and the timeout is set to 3 msec.
- OLT 320 sets the GPIO 324 then OLT 340 becomes the new standby OLT 354 .
- the OLT monitors the EPON network, but these conventional systems still need to solve the synchronization and security issues previously described.
- the embodiment currently being described provides a more general solution where the standby (redundant) OLT 320 does not monitor the EPON network (listen to upstream traffic) because the standby OLT does not know when to expect to receive data and thus the standby OLT cannot synchronize to received data and cannot learn the EPON network.
- the standby OLT 320 When the main OLT 340 triggers 348 the GPIO, the GPIO trigger is detected 322 by the standby OLT 320 . Note that when the standby OLT 320 is triggered and begins the process of becoming the new main OLT, the laser of the standby OLT is off. At any given time at most one OLT has a laser that is turned on, to facilitate correct multiplex control of the passive splitter 106 ( FIG. 2 ). The standby OLT 320 switchover process starts by setting the GPIO 324 . The host is notified 326 that the standby OLT 320 is operating as the main OLT, and then the OLT 320 waits 328 a timeout period, in one implementation 3 msec.
- This timeout period is necessary so the communications channels to the ONUs will be quiet, in other words in order to clear the line.
- the ONUs recognizes a period of time with no downstream traffic, stop transmitting, and transition to a lenient state.
- the OLT 320 turns on 330 the transmit laser.
- the RTTs are then calibrated 332 from the new main OLT 320 to the ONUs.
- One technique for measuring the RTT is to send GATE commands from the OLT 320 to the ONUs to recalibrate the RTT for each ONU.
- Another technique is to adapt the RTTs that were received from the host during the regular operation of synchronizing the databases from the main OLT to the standby OLT.
- the process of calibrating the RTT is comparable to pedal ning a dedicated discovery process for each ONU, where each ONU is already registered using GATE/REPORT messages.
- the OLT sends to each ONU a GATE for a REPORT message.
- the received REPORT message is used to calculate the RTT.
- the original standby OLT 320 starts operating 334 as the new main OLT, including starting operating the DBA. If applicable, the OLT 320 sends 336 IGMP queries in order to update the IGMP proxy table.
- the IGMP and MAC address tables cannot be copied from a main OLT to a standby OLT in real time. Therefore, after starting operation the new main OLT will learn missing information such as IGMP and MAC address table information.
- the host 310 receives 314 notification 350 from the main OLT 340 that the main OLT will no longer function as the main OLT or receives notification 326 from the standby OLT 320 that the standby OLT is operating as the main OLT.
- the host 310 notifies 316 the network switch 300 of the switchover from the main OLT 340 to the standby OLT 320 so the network switch can exchange ports 304 to stop routing data to the main OLT 340 and start routing data to the standby OLT 320 .
- Statistics counters are maintained by each OLT to facilitate monitoring and troubleshooting. Statistics can be read from both the active main OLT and from the old main/standby OLT. Synchronization of statistics between OLTs is not necessary.
- a host is reading statistics from a first OLT that is operating as the active main OLT. A switchover occurs, and a second OLT is now operating as the main OLT. The host now reads statistics from the second OLT. If the host can still communicate with the first OLT, the host can also read (any remaining) statistics from the first OLT. This feature can be used when a failure occurs, or to facilitate manual switchover, such as in a case where maintenance or an upgrade to an OLT is needed. The time required for switchover can be calculated as follows: The times given are for the worst-case scenario:
- an ONU generally operates in a normal state, shown as block 400 . Details of normal state operation for a conventional ONU are described in the IEEE 802.3 specification. As described elsewhere in this document, in a normal state, an ONU is configured to monitor an associated communication channel. Depending on the application, monitoring can be implemented using various techniques. In an optional implementation, the ONU monitors the associated communication channel at the physical layer, shown in block 402 , for example to see if the ONU is receiving a signal. An alternative to physical layer monitoring, block 402 , is a preferred implementation shown in block 404 where the ONU monitors if GATE commands are being received, and where lack of GATE commands indicates that the monitored communications channel is quiet.
- An ONU in a lenient state allows the ONU to update the timestamp variable on the ONU without deregistering, facilitating a fast recovery from failure. Accepting only a one-time adjustment of the ONU timestamp counter provides protection against replay attacks on the ONU.
- Another protection against replay attacks that can be implemented in a communications network is to assign each packet a security counter.
- One non-limiting example is IEEE standard 802.1 AE where each packet gets a security counter which increments from one packet to the next transmitted packet. When a conventional ONU receives a packet with a wrong security counter, the ONU will ignore the packet.
- An ONU in a lenient state allows the ONU to accept one time a new security counter, or in other words accept a bad security counter. Accepting a new security counter allows the ONU to continue receiving packets, preventing a quiet state, avoiding deregistering, and facilitating a fast recovery from failure.
- Implementation of the above-described lenient state features can vary depending on the specific application.
- a non-limiting example of implementing the feature of one-time acceptance of a new security counter uses the replayProtect variable in IEEE standard 802.1AE.
- the replayProtect variable When an ONU enters the lenient state, the replayProtect variable is turned off, allowing the ONU to accept a new security counter. When a packet is received by the ONU and a new security counter has been accepted (as well as the ONU performing the other above-described functions), the ONU transitions to the normal state and the replayProtect variable is turned on. Based on this description one familiar with the art will be able to implement appropriate techniques.
- the new security key is sent to the associated OLT, the ONU starts using the new security key, and the ONU and the OLT retain the security key that was being used, now called the old security key. This retaining of the old security key occurs every time a new security key is issued.
- a first security key is being used by an ONU.
- the new security key is issued, the first security key is no longer used for encryption and decryption, but is retained by the ONU and by the OLT as the old security key.
- the first security key is discarded, the second security key becomes the old security key, and the third (new) security key is used for encryption and decryption of transmitted and received packets.
- the ONU does not receive data for a minimum length of time, the ONU transitions to the lenient state, as described above. In the lenient state both the second (old) and third (new) security keys are used to try to decrypt packets. If the ONU transmits data while in the lenient state, the ONU uses the old security key to encrypt the data to be transmitted because the new OLT may not yet have the new security key that was being used between the ONU and the old OLT.
- the new OLT will be able to decrypt with the old security key. If even using both the old and new keys, no recognizable packets are received, after a given length of time the ONU transitions to the deregistered state. If a packet is successfully decrypted and recognized using the second security key (in our continuing example the old security key), the ONU's timestamp is updated with the timestamp of the received packet, the ONU transitions to the normal state, and the second key is used by the ONU to encrypt/transmit and decrypt/receive data. Now in the normal state, when a new fourth security key is issued the third security key is not retained, the second security key is retained as the old security key and the fourth security key becomes the new security key (the third security key is discarded).
- the fault tolerance of the system can be improved by implementing redundancy in the various components and communication channels.
- communications channel 234 is implemented with redundant connections to each OLT and between the hosts.
- redundant network switches 220 and redundant associated communication channels 230 A, 230 B, 232 can be implemented. Based on this description and the requirements of the application, one skilled in the art will be able to implement a system providing the required fault tolerance.
- FIG. 5 a diagram of an example of an N:1 redundant EPON network, three EPON networks are shown with a single standby OLT.
- EPON network-1, 500 is similar to the description of FIG. 2 .
- OLT-1 200 A is deployed at a first location connected via communication channel 104 A to passive optical splitter 106 .
- Communication channels 507 operationally connect the passive optical splitter 106 to one or more associated ONUs 214 A, 214 B, and 214 N.
- the hosts, network switches, communication channels, and other components are not drawn in FIG. 5 for each EPON.
- EPON network-2 530 includes OLT-3 532 connected via communication channel 534 to passive optical splitter 536 .
- Communication channels 537 operationally connect the passive optical splitter 536 to one or more associated ONUs 538 A, 538 B, and 538 N.
- EPON network- 3 540 includes OLT-N 542 connected via communication channel 544 to passive optical splitter 546 .
- Communication channel 547 operationally connects the passive optical splitter 546 to one or more associated ONUs 548 A, 5488 , and 548 N.
- a common standby OLT-2 200 B is deployed at a second location and connected to passive splitters 106 , 536 , and 546 via communication channels 1048 , 534 B, and 544 B, respectively.
- OLT-2 200 B is connected to OLT-1 200 A, OLT-3 532 , and OLT-N 542 via communication channels 236 , 539 , and 549 , respectively, to provide the above-described GPIO functionality.
- the components can be deployed co-located or in physically diverse locations. Based on the above description and example N:1 network, one skilled in the art will be able to implement other N: 1 networks.
Abstract
A system for redundancy in Ethernet passive optical networks (EPONs) facilitates fast recovery from failure (less than 50 msec), path redundancy of the fiber optic network, and location redundancy of the OLTs. An optical networking unit (ONU) in a normal state monitors input communications, and when the input communications are quiet for a predetermined minimum length of time, the ONU transitions to a lenient state in which: the ONU accepts old and new security keys; upon receiving a packet: the ONU updates an ONU timestamp based on the packet's timestamp; and the ONU transitions to the normal state of operation. While the ONU is in the lenient state if a packet is not received for a predetermined given length of time the ONU transitions to a deregistered state. In this system, main and standby OLTs do not require synchronization of security parameters or synchronization for differences in fiber lengths.
Description
- This application is a continuation of U.S. patent application Ser. No. 12/697,253 filed 31 Jan. 2010, which is incorporated by reference.
- The present embodiment generally relates to the field of communications networks, and in particular, concerns a system for redundancy in Ethernet passive optical networks.
- Background information on technology and terms can be found in the IEEE 802.3 specification (particularly clauses 64 and 65 for 1 gigabit [1 G] and the to-be-released clauses 76 and 77 for 10 gigabit [10 G]). The need for redundancy in networks, and in particular in Ethernet passive optical networks (EPONs) is known in the art. Referring to
FIG. 1 , a diagram of an EPON network, an OLT (optical line transmission equipment of the network provider) (100, 101) communicates over a fiber optic network (104A, 104B, and 107A, 107S, 107N) with ONUs (optical networking units associated with a user) 108A, 108B, 108N. Apassive splitter 106 facilitates either OLT-0 100 or OLT-1 101 communicating with the ONUs 108A, 108B, 108N. Ahost 124 provides configuration information, command, control, and monitoring of the OLTs (100, 101). A variety of architectures can be used to provide redundancy of the fiber optic network and the EPON communications standard includes features to facilitate recovery from a network failure. Redundancy in an EPON system facilitates continuity of service including switchover from a first OLT (sometimes called the “main”) to a second OLT (sometimes called the “standby”) and/or from a first network to a second network. This switchover can include alternative OLTs and fiber optic cables. - US publication 2008/0037981 A1 to Toshiaki Mukojima for Optical communication system with N+1 redundancy teaches a system for two or more active interfaces, each controlling the transmission and reception of optical signals between a communication network and one or more subscriber terminals. The control information used by all the active interfaces is stored in a common memory accessible to all of the interfaces. If a fault is detected in an active interface, a standby interface extracts the control information of the faulty interface from the memory and sends the control information to the standby interface. The interfaces are necessarily co-located because the control information includes the RTT (round trip time) data. Also required is an optical switching apparatus that switches data transmission paths among the network, the active and standby interfaces, and the subscriber terminals.
- US publication 2008/0025724 A1 to Hirokazu Ozaki for PON system, station side device and redundant method used for the same teaches a system including a plurality of PONs, a plurality of ONUs, and a station side device including a plurality of PON interface sections using one or more optical switch devices to provide 1:N redundancy of the PONs by cascading the N optical switch modules. The PON interface sections are co-located in the station side device and it is assumed that a value of an adjusted optical channel length inside the station side device is set.
- US publication 2007/0268818 A1 to Youichi Sugihara for Station side transmission unit, operation control method for station side transmission unit, and optical networking using station side transmission unit teaches a method for redundant units closing and releasing closure of their interfaces to switch between active units for a PON network. While this method provides physical layer switching, this method does not address the service or operational aspects of network failure and recovery.
- In the context of this document a failure includes a problem with the OLT (100, 101), such as an internal logic failure or general power failure, or a problem with the fiber optic network, such as a break in the fiber
optic cable first node 106 on the fiber optic network. Many conventional redundancy techniques provide switchover from a main OLT to a standby OLT, and after a delay, service is restored to the ONU in accordance with the established specifications. OLT redundancy in EPON systems is relatively simple when the switchover time can be relatively long, approximately 1 second. In this case, the ONUs deregister, the standby OLT comes online, all the ONUs re-register, and service is eventually restored to all of the ONUs. - An ONU will deregister under certain conditions, including:
- 1. A long time, normally greater than 300 msec, has passed without the ONU receiving any downstream GATE packets.
- 2. There is a timestamp drift of greater than 16 TQ (256 nsec) in the timestamp received from the OLT by the ONU (1 TQ, or Time Quanta is 16 nsec). A timestamp drift can occur when switching from a first OLT using a first clock to a second OLT using a second clock, where the clocks are not synchronized. Even with synchronized clocks, a difference in the lengths of the fiber optic cable from each of the OLTs to the first node on the fiber optic network is a source of timestamp drift.
- 3. A new OLT not having the current security key for an ONU. When an OLT sends data using a security key that an ONU is no longer using, the ONU will not recognize the data. Because the ONU does not recognize the data, the ONU will be in a state of not receiving data, and after a given time has elapsed without the ONU receiving data, the ONU will deregister. In a case where the control packets are not encrypted, the ONU will still not be able to recognize the data, and will be in a state of not receiving data. In the context of this document, data is used as a general reference to all information communicated, unless otherwise specified. In a case where the data that the system secures is only the data packets, but the control packets are sent unencrypted, the ONU will similarly be in a state of not receiving data.
- When an ONU deregisters, the conventional time to recover is approximately 1 second and typically between 300 msec and 1500 msec. This conventional time to recover is no longer sufficient for new applications being deployed and applications planned for future deployment. Application providers are now demanding a time to recover of less than 50 msec in order to provide new applications and a higher level of service to users. To achieve recovery within this timeframe, the ONUs cannot be allowed to deregister when switching to a redundant OLT.
- Conventional redundancy implementations typically co-locate the redundant OLTs in order to share common resources, synchronize OLTs, or minimize differences in the length of fiber optic cables. A popular implementation of OLTs is as a printed circuit board, with multiple OLTs deployed in a single chassis at a network provider's facility and a single host controlling the OLTs. However, in a case where each OLT has a clock generator for the OLT's timestamp counter, the timestamp counters are not synchronized. The timestamp counters advance every 16 nsec (1 TQ), and synchronization requires at least 2 wires (1 wire for clock and 1 wire for reset) connected between the OLTs. This case is a difficult implementation with strict physical limitations. There are variances in the round trip delay (RTT) due to the different lengths of the fiber optic cable from each of the OLTs to the first node on the fiber optic network. In this case, synchronizing the RTT is a manual process that again has strict physical limitations. In addition, co-locating OLTs and hosts exposes the system to failure due to lack of location redundancy. In the context of this description, location redundancy refers to deploying portions of the system at physically diverse locations. In one non-limiting example of location non-redundancy, a main and a standby OLT are both located in the same room. A natural disaster, such as flooding, an accident, such as a fire, or power shutdown for routine maintenance affect both the main and standby OLTs resulting in a failure of the EPON.
- Conventional techniques for synchronizing security keys include maintaining a real-time database of the current security key for each ONU. As stated earlier, if a new OLT does not have the current security key for an ONU then the ONU will not recognize the data being sent from the new OLT, potentially causing the ONU to deregister and the data to be discarded. Synchronizing security keys at real time between two OLTs is a complex technique, possibly requiring extensive hardware wiring between OLTs.
- It is thus desirable to have a system for EPON redundancy that provides recovery in less than 50 msec. It is further desirable to have redundancy of multiple layers of the EPON, including network provider OLT (equipment redundancy), the fiber optic network (path redundancy), and physical location (location redundancy).
- A system for redundancy in Ethernet passive optical networks (EPONs) includes: at least one optical networking unit (ONU), wherein when the ONU is in a normal state the ONU monitors input communications, and wherein when the input communications are quiet for a predetermined minimum length of time the ONU transitions to a lenient state in which: the ONU accepts old and new security keys; upon receiving a packet: the ONU updates an ONU timestamp based on said packet's timestamp; and the ONU transitions to said normal state of operation; and wherein while the ONU is in the lenient state, if a packet is not received for a predetermined given length of time the ONU transitions to a deregistered state.
- In an alternate implementation, upon receiving the packet the ONU further updates an ONU security counter based on the packet's security counter.
- In an alternate implementation, the system further includes at least two optical line transmission equipment (OLT) wherein a first OLT is deployed at a first location and a second OLT is deployed at a second location: and a passive optical splitter for exchanging packets between the OLTs and the at least one ONU.
- In an alternate implementation, the OLTs are configured to transmit a switchover signal and the OLTs are configured to receive a switchover signal from other OLTs. In another alternate implementation, when an OLT receives the switchover signal, the OLT executes a switchover process. In another alternate implementation, the OLT is configured such that after the switchover process IGMP and MAC address table information is determined.
- In another alternate implementation, the system further includes one or more hosts configured to transmit a switchover signal, and wherein the OLTs are configured to receive said switchover signal from the one or more hosts. In another alternate implementation, when an OLT receives said switchover signal, the OLT executes a switchover process. In another alternate implementation, the system further includes one or more network switches operationally connected to the OLTs, the one or more network switches configured to interface between the OLTs and a network. In another alternate implementation, the OLTs are configured to maintain independently statistic counters for monitoring and troubleshooting.
- In an alternate implementation, the system is configured to switch from using the first OLT to using the second OLT in less than 50 msec.
- In an alternate implementation, system is configured to recalculate the round trip time (RTT) after OLT switchover while the at least one ONU is still registered.
- The embodiment is herein described, by way of example only, with reference to the accompanying drawings, wherein:
-
FIG. 1 , a diagram of an EPON network -
FIG. 2 , a diagram of a redundant EPON network -
FIG. 3 , a timing flow diagram of the OLT switchover process -
FIG. 4 , a flow diagram of ONU lenient mode operation -
FIG. 5 , a diagram of an example of an N:1 redundant EPON network. - The principles and operation of the system according to the present embodiment may be better understood with reference to the drawings and the accompanying description. The present embodiment is a system for redundancy in Ethernet passive optical networks (EPONs). The system facilitates fast recovery from failure, specifically with a recovery time of less than 50 msec. The system also facilitates path redundancy of the fiber optic network. The system also facilitates implementation with location redundancy of the OLTs, which facilitates greater path redundancy of the fiber optic network. Security information for the standby OLT can be updated asynchronously from the main OLT. Note that this system can be used to provide N:1 redundancy. For clarity, the description first describes an implementation of 2:1 redundancy and then describes how this implementation can be extended to provide an N:1 implementation.
- Referring now to the drawings,
FIG. 2 , a diagram of a redundant EPON network, a first optical line transmission equipment (OLT) 200A is deployed at a first location and asecond OLT 200B is deployed at a second location. Deploying each OLT at different physical locations facilitates location redundancy. The specific application and desired level of location redundancy will provide requirements for determining what locations will be sufficient to satisfy the location redundancy requirements. In one implementation, the OLTs are deployed in the same room, or even in the same chassis, with each OLT having independent power supplies and network connections. In another implementation, each OLT is deployed in a different building. The location redundancy configuration of OLTs will be limited by the operating parameters of the EPON specification and deployment realities. The OLTs are configured to send and receive packets overfirst communication channels optical splitter 106.Second communication channels optical splitter 106 to one or more associated optical networking units (ONUs) 214A, 21413, and 214N, respectively. In the context of this description, an active communication channel is a communication channel from which packets are received and to which packets can be transmitted between given devices on the communications channel. In the context of this description a quiet communications channel is a communications channel which is not communicating packets between given devices on the communications channel. A quiet communications channel can physically have signals being carried on the communications channel and reaching the physical input of a device on the network, but if the signal is not intended for the receiving device, or if the receiving device is not able to process the signal, the communications channel is considered quiet for that device. A non-limiting example of a quiet channel is the case where packets are being transmitted from an OLT to an ONU, but the ONU does not have the correct security key to decrypt the received packets. In this case, the ONU is not receiving usable data and the input communications channel is considered quiet. - In a normal state, the ONUs are configured to monitor an associated second communication channel. When an associated second communication channel is quiet for a minimum length of time, a conventional ONU transitions from a normal state to a deregistered state, requiring approximately 1000 msec to reregister during which application data is not transmitted or received. In one implementation, to avoid deregistration the ONUs are configured with an innovative “lenient state”. When an associated communication channel is quiet for a minimum length of time, the ONU transitions from a normal state to a lenient state. In the lenient state, the ONU starts accepting both the old (the previously used) security key and the current security key for decrypting received packets, as described below. The old security key is used for encrypting and transmitting packets. This lenient state allows the ONU to receive a packet, the ONU's timestamp is updated based on the packet's timestamp, and the ONU's security counter is updated based on the packet's security counter. After a packet is received in the lenient state, the ONU transitions to a normal state of operation. When the ONU is in a lenient state and a packet is not received for a given length of time, the ONU transitions to a deregistered state. Using a lenient state reduces the switchover time from an average of 1000 msec to a worst case of 21 msec, as will be described below.
-
Hosts communication channel 234.Network switch 220 is connected to OLT-1 200A and OLT-2 200B bycommunication channels network switch 220 provides connectivity viacommunications channel 232 to thecore network 222. Communications channel 236 between OLT-1 200A and OLT-2 200B is for general-purpose input/output pins (GPIO) connection and in the context of this document is referred to as the GPIO connection. GPIO output pins from OLT-1 200A connect to GPIO input pins on OLT-2 200B. Similarly, GPIO output pins from OLT-2 200B connect to GPIO input pins on OLT-1 200A. In an implementation where the OLTs are relatively close, a cable can be used to connect directly to the OLTs' GPIO. In an implementation where the OLTs are deployed in a physically diverse configuration, the GPIO communication channel can be remoted as appropriate to the application. Techniques for remoting communicating channels are known in the art, and based on this description one familiar with the art will be able to implement an appropriate technique. The GPIO connection is configured to transmit a switchover signal from the main OLT-1 200A to the standby OLT-2 200B. When OLT-2 200B receives a switchover signal, the switchover process begins for OLT-2 200B. One preferred implementation for transmitting a switchover signal is by lowering the main OLT's GPIO to ‘0’. This can be done by setting the GPIO to input mode where the output gets a ‘high Z’ value and is pulled down by the pull-down resistor. In this preferred implementation of transmitting a switchover signal, if the main OLT signals a low level or if there is a power failure of the main OLT, the standby OLT receives a switchover signal. This allows the standby OLT to recognize a power failure of the main OLT. Note that as described elsewhere, after the switchover process is complete OLT-2 200E has become the main OLT and. OLT-1 200A has become the standby OLT in preparation for future switchovers. - Fiber protection switching, also known as OLT switchover, can be initiated by a variety of circumstances including:
-
- The main OLT has detected a problem with the fiber communications channel;
- The standby OLT recognizes a power failure of the main OLT; and
- A switchover command is given to the OLTs.
- The standby and main OLTs can be deployed in a variety of configurations, including, but not limited to, each OLT located in a different physical location, the OLT's co-located but deployed in separate chassis, or OLT printed circuit boards (PCBs) deployed in the same chassis. In a case where the OLTs are implemented on separate printed circuit boards in the same chassis, there is no need to synchronize clock and reset signals between the two boards. Each OLT has an associated fiber for communications with the EPON. The choice of initially which OLT is the main OLT and which OLT is the standby OLT is dependent on the specifics of the application and deployment requirements. Once in use, either unit can function as the main OLT and the other unit will be the standby OLT. After a switchover is completed, the former main OLT becomes the standby OLT and the former standby OLT operates as the main OLT. It is not necessary to maintain a specific OLT permanently in a specific role as a main or standby OLT.
- In this description, we arbitrarily choose OLT-1 200A, as the initially main OLT and OLT-2 200B as the initially standby OLT to assist in describing the OLT switchover. The standby OLT-2 200B is passive, not initiating transactions on
communications channel 104B to the EPON or to the network switch viacommunications channel 230B. Similarly, one of thehosts 224A is considered the main host and provides configuration information, command, control, and monitoring of the OLTs. Information maintained by the host includes which ONUs are registered and dynamic bandwidth allocation (DBA) order of the ONUs, address table, IGMP proxy table, ONU networking settings, security status (enabled/keys) of each ONU, and static OLT configurations such as classification rules. Themain host 224A actively updates and is updated by the main OLT-1 200A. Themain host 224A updates the standby OLT-2 200B and can update thestandby host 224B. - Referring to
FIG. 3 , a timing flow diagram of the OLT switchover process,column 300 is for actions of the switch,column 310 the host,column 320 the initial standby OLT, andcolumn 340 the initial main OLT. The flow down the page isrelative time 302. Note that lengths of time described in this document are preferable, but timeouts and waiting periods are configurable, and the system can be configure with specific lengths of time appropriate to the application. The switchover process can be initiated in a variety of ways, including, but not limited to: thehost 310 issuing astart switchover command 312 to themain OLT 340, the main OLT detecting aproblem 342 with the fiber communications channel, the main OLT not detecting any upstream traffic, or by detecting a problem with the fiber optic network (for example via fiber monitoring), in which cases themain OLT 340 stops transmitting 344, by turning off the transmit laser. - After transmission has stopped 344, the
main OLT 340 triggers 348 the GPIO. The switchover process can also be initiated when themain OLT 340 has apower failure 346, in which case due to loss of power the GPIO is triggered 348 (and the laser turns off due to the power failure). In one implementation, the GPIO is triggered according to the pulldown resistor technique described above. In one case, the switchover command is given to the main OLT to take the main OLT out of use and perform maintenance on the main OLT. In another case, if the main OLT is malfunctioning, the OLT still has power and the GPIO will not be pulled down. One non-limiting example of a malfunctioning OLT is where there is a problem with a processor on the OLT (commonly known as a “stuck” or “hung” CPU). Another non-limiting example of a malfunctioning OLT is where there is a problem with the OLT PON receive line, resulting in a quiet communications channel. For this case, a watchdog process on the host monitors the main OLT, and if the main OLT does not respond to the watchdog process then the host initiates a start switchover command to the standby OLT to start the switchover process. - Continuing with the switchover process for the
main OLT 340 in the case where there is still power to the main OLT, after the GPIO is triggered the main OLT notifies 350 thehost 310 that the main OLT will no longer function as the main OLT. TheOLT 340 waits 352 a timeout period to allow theoriginal standby OLT 320 to start functioning as the new main OLT and then theOLT 340 becomes thenew standby OLT 354. The timeout period should preferably be at least twice the dynamic bandwidth allocation (DBA) cycle to assure that none of the ONUs has to transmit any more upstream packets. In one non-limiting example, the DBA cycle is 0.5 msec and the timeout is set to 3 msec. Alternatively, when theOLT 320 sets theGPIO 324 thenOLT 340 becomes thenew standby OLT 354. - In some conventional systems, the OLT monitors the EPON network, but these conventional systems still need to solve the synchronization and security issues previously described. The embodiment currently being described provides a more general solution where the standby (redundant)
OLT 320 does not monitor the EPON network (listen to upstream traffic) because the standby OLT does not know when to expect to receive data and thus the standby OLT cannot synchronize to received data and cannot learn the EPON network. - When the
main OLT 340 triggers 348 the GPIO, the GPIO trigger is detected 322 by thestandby OLT 320. Note that when thestandby OLT 320 is triggered and begins the process of becoming the new main OLT, the laser of the standby OLT is off. At any given time at most one OLT has a laser that is turned on, to facilitate correct multiplex control of the passive splitter 106 (FIG. 2 ). Thestandby OLT 320 switchover process starts by setting theGPIO 324. The host is notified 326 that thestandby OLT 320 is operating as the main OLT, and then theOLT 320 waits 328 a timeout period, in oneimplementation 3 msec. This timeout period is necessary so the communications channels to the ONUs will be quiet, in other words in order to clear the line. The ONUs recognizes a period of time with no downstream traffic, stop transmitting, and transition to a lenient state. After waiting 328 a timeout period, theOLT 320 turns on 330 the transmit laser. The RTTs are then calibrated 332 from the newmain OLT 320 to the ONUs. One technique for measuring the RTT is to send GATE commands from theOLT 320 to the ONUs to recalibrate the RTT for each ONU. Another technique is to adapt the RTTs that were received from the host during the regular operation of synchronizing the databases from the main OLT to the standby OLT. The process of calibrating the RTT is comparable to pedal ning a dedicated discovery process for each ONU, where each ONU is already registered using GATE/REPORT messages. The OLT sends to each ONU a GATE for a REPORT message. The received REPORT message is used to calculate the RTT. - The
original standby OLT 320 starts operating 334 as the new main OLT, including starting operating the DBA. If applicable, theOLT 320 sends 336 IGMP queries in order to update the IGMP proxy table. The IGMP and MAC address tables cannot be copied from a main OLT to a standby OLT in real time. Therefore, after starting operation the new main OLT will learn missing information such as IGMP and MAC address table information. - During the switchover process, the
host 310 receives 314notification 350 from themain OLT 340 that the main OLT will no longer function as the main OLT or receivesnotification 326 from thestandby OLT 320 that the standby OLT is operating as the main OLT. Thehost 310 notifies 316 thenetwork switch 300 of the switchover from themain OLT 340 to thestandby OLT 320 so the network switch can exchangeports 304 to stop routing data to themain OLT 340 and start routing data to thestandby OLT 320. - Statistics counters are maintained by each OLT to facilitate monitoring and troubleshooting. Statistics can be read from both the active main OLT and from the old main/standby OLT. Synchronization of statistics between OLTs is not necessary. In one non-limiting example, a host is reading statistics from a first OLT that is operating as the active main OLT. A switchover occurs, and a second OLT is now operating as the main OLT. The host now reads statistics from the second OLT. If the host can still communicate with the first OLT, the host can also read (any remaining) statistics from the first OLT. This feature can be used when a failure occurs, or to facilitate manual switchover, such as in a case where maintenance or an upgrade to an OLT is needed. The time required for switchover can be calculated as follows: The times given are for the worst-case scenario:
-
- The main OLT does not respond to a watchdog process: 3 msec
- Trigger the GPIO switchover signal: 1 msec
- Standby (new main) OLT waits for the EPON to go into the lenient state: 3 msec
- RTT calculations for 64 ONUs: 14 msec
- for a total delay of 21 msec, which is below the desired 50 msec limit on switchover time.
- For reference, the required time for the round trip time (RTT) delay calculation for 64 ONUs is as follows:
-
1 meter of fiber takes 1/(speed of light in fiber)=1 m/(200000000 m/s)=5 nsec -
The maximum fiber length from OLT to ONU=20 km which takes 20 km*5 nsec/m=100 uSec. -
- The round trip time from an OLT to an ONU back to the OLT is:
-
[100 usec from an OLT to an ONU]+[20 usec for GATE processing in the ONU which is 1250 TQ from GATE to REPORT]+[100 usec from the ONU back to the OLT]=220 usec -
64 ONUs*220 usec=14 msec - Referring to
FIG. 4 , a flow diagram of ONU lenient mode operation, an ONU generally operates in a normal state, shown asblock 400. Details of normal state operation for a conventional ONU are described in the IEEE 802.3 specification. As described elsewhere in this document, in a normal state, an ONU is configured to monitor an associated communication channel. Depending on the application, monitoring can be implemented using various techniques. In an optional implementation, the ONU monitors the associated communication channel at the physical layer, shown inblock 402, for example to see if the ONU is receiving a signal. An alternative to physical layer monitoring, block 402, is a preferred implementation shown inblock 404 where the ONU monitors if GATE commands are being received, and where lack of GATE commands indicates that the monitored communications channel is quiet. When an associated communication channel is quiet for a minimum length of time, the ONU transitions from a normal state to a lenient state, block 406. If the ONU does not receive a packet from the OLT within a given length of time while in the lenient state, the ONU has timed-out, shown inblock 408, of the lenient state and the ONU transitions to the deregistered state, block 420. After deregistering, block 420, the ONU goes through conventional reregistration, block 422, and transitions to the normal state of operation. - While the ONU is in the lenient state, block 408, the ONU accepts both old and new security keys, shown in
block 410, and described below. In other words, both the old and current security keys are used to try to decode received packets. Note that when the ONU enters the lenient state the ONU starts accepting old and new keys, shown inblock 410, after which the GNU continues to accept the same old and new keys. In an alternate implementation, after switchover the OLT sends unsecured packets to the GNU, and the ONU while in lenient mode accepts unsecured packets and transmits unsecured packets, until a new security connection is established, in other words until a new security key has been exchanged. While the communications is quiet and the ONU has not yet received a packet, block 412, the GNU remains in the lenient state, block 406. If the ONU receives a packet, the ONU's timestamp is updated, block 414, based on the packet's timestamp, the GNU accepts the security counter, block 415, of the packet, and the ONU transitions to the normal state of operation, block 400. As described previously, if there is a timestamp drift of greater than 16 TQ (256 nsec) in the timestamp received from an OLT by an ONU, conventional ONUs will deregister. An ONU in a lenient state allows the ONU to update the timestamp variable on the ONU without deregistering, facilitating a fast recovery from failure. Accepting only a one-time adjustment of the ONU timestamp counter provides protection against replay attacks on the ONU. Another protection against replay attacks that can be implemented in a communications network is to assign each packet a security counter. One non-limiting example is IEEE standard 802.1 AE where each packet gets a security counter which increments from one packet to the next transmitted packet. When a conventional ONU receives a packet with a wrong security counter, the ONU will ignore the packet. As long as the received packets continue to have a security counter that is not acceptable, the ONU will continue to ignore the packets, resulting in a quiet state for the ONU and eventually deregistration of the ONU. An ONU in a lenient state allows the ONU to accept one time a new security counter, or in other words accept a bad security counter. Accepting a new security counter allows the ONU to continue receiving packets, preventing a quiet state, avoiding deregistering, and facilitating a fast recovery from failure. Implementation of the above-described lenient state features can vary depending on the specific application. A non-limiting example of implementing the feature of one-time acceptance of a new security counter uses the replayProtect variable in IEEE standard 802.1AE. When an ONU enters the lenient state, the replayProtect variable is turned off, allowing the ONU to accept a new security counter. When a packet is received by the ONU and a new security counter has been accepted (as well as the ONU performing the other above-described functions), the ONU transitions to the normal state and the replayProtect variable is turned on. Based on this description one familiar with the art will be able to implement appropriate techniques. - To implement the security feature of the lenient state, when an ONU is in the normal state and the ONU issues a new security key, the new security key is sent to the associated OLT, the ONU starts using the new security key, and the ONU and the OLT retain the security key that was being used, now called the old security key. This retaining of the old security key occurs every time a new security key is issued. In a non-limiting example, a first security key is being used by an ONU. When a second security key, also called the new security key is issued, the first security key is no longer used for encryption and decryption, but is retained by the ONU and by the OLT as the old security key. When a third security key is issued, the first security key is discarded, the second security key becomes the old security key, and the third (new) security key is used for encryption and decryption of transmitted and received packets. If the ONU does not receive data for a minimum length of time, the ONU transitions to the lenient state, as described above. In the lenient state both the second (old) and third (new) security keys are used to try to decrypt packets. If the ONU transmits data while in the lenient state, the ONU uses the old security key to encrypt the data to be transmitted because the new OLT may not yet have the new security key that was being used between the ONU and the old OLT. Note that if the new OLT does have the new security key, the new OLT will be able to decrypt with the old security key. If even using both the old and new keys, no recognizable packets are received, after a given length of time the ONU transitions to the deregistered state. If a packet is successfully decrypted and recognized using the second security key (in our continuing example the old security key), the ONU's timestamp is updated with the timestamp of the received packet, the ONU transitions to the normal state, and the second key is used by the ONU to encrypt/transmit and decrypt/receive data. Now in the normal state, when a new fourth security key is issued the third security key is not retained, the second security key is retained as the old security key and the fourth security key becomes the new security key (the third security key is discarded).
- Referring back to
FIG. 2 , the fault tolerance of the system can be improved by implementing redundancy in the various components and communication channels. In one implementation, instead ofcommunications channel 234 being implemented as a single shared network,communications channel 234 is implemented with redundant connections to each OLT and between the hosts. In other implementations redundant network switches 220 and redundant associatedcommunication channels - As previously mentioned, the above description of an implementation of 2:1 redundancy can be extended to provide an N:1 implementation. Referring to
FIG. 5 , a diagram of an example of an N:1 redundant EPON network, three EPON networks are shown with a single standby OLT. EPON network-1, 500 is similar to the description ofFIG. 2 . OLT-1 200A is deployed at a first location connected viacommunication channel 104A to passiveoptical splitter 106.Communication channels 507 operationally connect the passiveoptical splitter 106 to one or more associatedONUs FIG. 5 for each EPON. - Similar to EPON network-1, EPON network-2 530 includes OLT-3 532 connected via communication channel 534 to passive
optical splitter 536.Communication channels 537 operationally connect the passiveoptical splitter 536 to one or more associatedONUs N 542 connected via communication channel 544 to passiveoptical splitter 546.Communication channel 547 operationally connects the passiveoptical splitter 546 to one or more associatedONUs - A common standby OLT-2 200B is deployed at a second location and connected to
passive splitters communication channels N 542 viacommunication channels FIG. 2 , the components can be deployed co-located or in physically diverse locations. Based on the above description and example N:1 network, one skilled in the art will be able to implement other N:1 networks. - It will be appreciated that the above descriptions are intended only to serve as examples, and that many other embodiments are possible within the scope of the present invention as defined in the appended claims.
Claims (16)
1. A system for redundancy in Ethernet passive optical networks (EPONs) comprising:
(a) at least two optical line transmission equipment (OLT) wherein a first OLT is deployed at a first location and a second OLT is deployed at a second location;
(b) at least one optical networking unit (ONU); and
(c) a passive optical splitter for exchanging packets between the OLTs and said at least one ONU,
wherein the OLTs are configured to transmit a switchover signal and the OLTs are configured to receive said switchover signal from other OLTs.
2. The system of claim 1 wherein when an OLT receives said switchover signal, said OLT executes a switchover process.
3. The system of claim 2 wherein said. OLT is configured such that after said switchover process IGMP and MAC address table information is determined.
4. The system of claim 1 further comprising one or more hosts configured to transmit a switchover signal, and wherein the OLTs are configured to receive a switchover signal from said one or more hosts.
5. The system of claim 4 wherein when an OLT receives said switchover signal, said OLT executes a switchover process.
6. The system of claim 1 further comprising one or more network switches operationally connected to the OLTs, said one or more network switches configured to interface between the OLTs and a network.
7. The system of claim 1 wherein the OLTs are configured to maintain independently statistic counters for monitoring and troubleshooting.
8. The system of claim 1 wherein said system is configured to switch from using said first OLT to using said second OLT in less than 50 msec (milliseconds).
9. A system for redundancy in Ethernet passive optical networks (EPONs) comprising:
(a) at least two optical line transmission equipment (OLT) wherein a first OLT is deployed at a first location and a second OLT is deployed at a second location;
(b) at least one optical networking unit (ONU); and
(c) a passive optical splitter for exchanging packets between the OLTs and the at least one ONU.
wherein said system is configured to recalculate the round trip time (RTT) after OLT switchover while said at least one ONU is still registered.
10. The system of claim 9 wherein the OLTs are configured to transmit a switchover signal and the OLTs are configured to receive a switchover signal from other OLTs.
11. The system of claim 10 wherein when an OLT receives said switchover signal, said OLT executes a switchover process.
12. The system of claim 11 wherein said OLT is configured such that after said switchover process IGMP and MAC address table information is determined.
13. The system of claim 9 further comprising one or more hosts configured to transmit a switchover signal, and wherein the OLTs are configured to receive a switchover signal from said one or more hosts.
14. The system of claim 13 wherein when an OLT receives said switchover signal, said OLT executes a switchover process.
15. The system of claim 9 further comprising one or more network switches operationally connected to the OLTs, said one or more network switches configured to interface between the OLTs and a network.
16. The system of claim 9 wherein the OLTs are configured to maintain independently statistic counters for monitoring and troubleshooting.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/794,834 US20130188946A1 (en) | 2010-01-31 | 2013-03-12 | System for redundancy in ethernet passive optical networks (epons) |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/697,253 US8422887B2 (en) | 2010-01-31 | 2010-01-31 | System for redundancy in Ethernet passive optical networks (EPONs) |
US13/794,834 US20130188946A1 (en) | 2010-01-31 | 2013-03-12 | System for redundancy in ethernet passive optical networks (epons) |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/697,253 Continuation US8422887B2 (en) | 2010-01-31 | 2010-01-31 | System for redundancy in Ethernet passive optical networks (EPONs) |
Publications (1)
Publication Number | Publication Date |
---|---|
US20130188946A1 true US20130188946A1 (en) | 2013-07-25 |
Family
ID=44341751
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/697,253 Active 2030-10-20 US8422887B2 (en) | 2010-01-31 | 2010-01-31 | System for redundancy in Ethernet passive optical networks (EPONs) |
US13/794,834 Abandoned US20130188946A1 (en) | 2010-01-31 | 2013-03-12 | System for redundancy in ethernet passive optical networks (epons) |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/697,253 Active 2030-10-20 US8422887B2 (en) | 2010-01-31 | 2010-01-31 | System for redundancy in Ethernet passive optical networks (EPONs) |
Country Status (3)
Country | Link |
---|---|
US (2) | US8422887B2 (en) |
JP (1) | JP5570026B2 (en) |
CN (1) | CN102256188B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140093232A1 (en) * | 2012-10-01 | 2014-04-03 | David Hood | Method for protection of multi-wavelength passive optical network |
WO2015041651A1 (en) * | 2013-09-19 | 2015-03-26 | Hewlett-Packard Development Company, L.P. | Software defined optical network |
Families Citing this family (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101453666B (en) * | 2007-12-07 | 2012-07-04 | 华为技术有限公司 | Main and standby link protection method, loop system and apparatus for passive optical network |
EP2615800B1 (en) * | 2010-12-08 | 2016-03-30 | Mitsubishi Electric Corporation | Communications method for optical communications system, optical communications system, secondary station device, control device and program |
JP5423701B2 (en) * | 2011-02-15 | 2014-02-19 | 住友電気工業株式会社 | Station side apparatus, control method, and control method of PON system |
CN103518381A (en) * | 2011-05-17 | 2014-01-15 | 瑞典爱立信有限公司 | Protection for fibre optic access networks |
EP2621194A1 (en) * | 2012-01-30 | 2013-07-31 | Mitsubishi Electric R&D Centre Europe B.V. | Transparent protection switching operation in a pon |
CN102740176B (en) * | 2012-06-28 | 2015-01-21 | 华为技术有限公司 | Communication method and communication system |
JP5988814B2 (en) * | 2012-10-03 | 2016-09-07 | 三菱電機株式会社 | Communication device, slave station device, control device, communication system, and communication control method |
US9351055B2 (en) | 2013-08-15 | 2016-05-24 | Mellanox Technologies Ltd. | High-reliability active optical cable (AOC) with redundant emitters |
JP5941024B2 (en) * | 2013-08-23 | 2016-06-29 | 日本電信電話株式会社 | Optical transmission system, subscriber termination device, and optical signal transmission device |
CN105792027B (en) * | 2014-12-22 | 2019-11-29 | 南京中兴新软件有限责任公司 | Passive optical network, controller and the method for communication |
US20160234582A1 (en) * | 2015-02-10 | 2016-08-11 | Daniel Ronald | Method and system for redundancy in a passive optical network |
CN105049233B (en) * | 2015-06-23 | 2018-02-13 | 上海斐讯数据通信技术有限公司 | OLT device configuration information display methods, device and OLT device |
US10148348B2 (en) * | 2015-07-17 | 2018-12-04 | Corning Optical Communications LLC | Optical-electrical interface devices and systems with optical communication pathway redundancy |
CN105978652B (en) * | 2016-04-27 | 2018-02-13 | 刘巍 | Equipment, system and method at synchronous pair of Redundant Ethernet |
WO2017192894A1 (en) * | 2016-05-04 | 2017-11-09 | Adtran, Inc. | Systems and methods for performing optical line terminal (olt) failover switches in optical networks |
JP6743494B2 (en) * | 2016-06-03 | 2020-08-19 | 富士通株式会社 | Transmission system, communication device, and path switching method |
CN108134631A (en) * | 2017-12-05 | 2018-06-08 | 无锡路通视信网络股份有限公司 | A kind of OLT loops and the method using the OLT loop transfer data |
WO2021176696A1 (en) * | 2020-03-06 | 2021-09-10 | 日本電信電話株式会社 | Communication system and olt system |
WO2021210078A1 (en) * | 2020-04-14 | 2021-10-21 | 日本電信電話株式会社 | Optical communication system and master station device |
WO2024069748A1 (en) * | 2022-09-27 | 2024-04-04 | 日本電信電話株式会社 | Communication device switching method and communication system |
CN116800334B (en) * | 2023-06-09 | 2024-03-26 | 国网吉林省电力有限公司长春供电公司 | Data synchronous transmission optimization method and system based on analog optical fiber communication |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020071148A1 (en) * | 2000-12-12 | 2002-06-13 | Purse Christopher M. | Optical communications network and nodes for forming such a network |
US20040208510A1 (en) * | 2002-03-18 | 2004-10-21 | Katsuichi Ohara | Optical transmission equipment |
US6975586B1 (en) * | 1999-03-30 | 2005-12-13 | Nec Corporation | Protection switching method and apparatus for passive optical network system |
US20070058973A1 (en) * | 2005-07-29 | 2007-03-15 | Keiji Tanaka | Optical termination system |
US20090067835A1 (en) * | 2007-09-10 | 2009-03-12 | Charles Chen | Method and apparatus for protection switching in passive optical network |
US20100098407A1 (en) * | 2008-10-21 | 2010-04-22 | Teknovus, Inc. | Method and system for protection switching in ethernet passive optical networks |
US20100166419A1 (en) * | 2008-12-30 | 2010-07-01 | Broadlight, Ltd. | Techniques for protecting passive optical networks |
US20110211837A1 (en) * | 2010-02-26 | 2011-09-01 | Hitachi, Ltd. | Method of Recovery From Sleep State of an Onu in a Pon System Capable of Power Saving |
Family Cites Families (26)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6317829B1 (en) * | 1998-06-19 | 2001-11-13 | Entrust Technologies Limited | Public key cryptography based security system to facilitate secure roaming of users |
JP4201430B2 (en) * | 1999-04-16 | 2008-12-24 | 富士通株式会社 | Optical subscriber line termination equipment |
US20020110245A1 (en) * | 2001-02-13 | 2002-08-15 | Dumitru Gruia | Method and system for synchronizing security keys in a point-to-multipoint passive optical network |
JP4366885B2 (en) * | 2001-05-24 | 2009-11-18 | 日本電気株式会社 | Optical communication network, optical communication node device, and fault location specifying method used therefor |
KR100523357B1 (en) * | 2003-07-09 | 2005-10-25 | 한국전자통신연구원 | Key management device and method for providing security service in epon |
KR100594133B1 (en) * | 2003-08-21 | 2006-06-28 | 삼성전자주식회사 | Code Division Multiple Passive Optical Network Using Error Correction Code |
KR100547724B1 (en) * | 2003-08-26 | 2006-01-31 | 삼성전자주식회사 | Passive optical subscriber network based on Gigabit Ethernet that can stably transmit data and data encryption method using same |
US7287175B2 (en) * | 2004-07-19 | 2007-10-23 | Tellabs Petaluma, Inc. | Optical network terminal with low-power sleep logic that substantially extends the life of the battery after the AC main power supply has been lost |
WO2006023015A1 (en) * | 2004-08-05 | 2006-03-02 | Optical Solutions, Inc. | Optical network terminal with low power hibernation |
US20060093356A1 (en) * | 2004-10-28 | 2006-05-04 | Vereen Jerry D | Optical network that detects and removes Rogue ONTS |
CN101218792A (en) * | 2005-07-15 | 2008-07-09 | 泰克诺沃斯公司 | Method and apparatus for facilitating asymmetric line rates in an Ethernet passive optical network |
JP4650168B2 (en) * | 2005-08-30 | 2011-03-16 | Kddi株式会社 | Optical termination device switching method |
CN100488207C (en) * | 2005-09-23 | 2009-05-13 | 华为技术有限公司 | Operation method for passive optic network user terminal |
JP2007311953A (en) * | 2006-05-17 | 2007-11-29 | Nec Corp | Station-building side transmission device and its operation control method, and optical network using it |
JP2007324853A (en) * | 2006-05-31 | 2007-12-13 | Oki Electric Ind Co Ltd | Optical communication system |
JP4992472B2 (en) * | 2006-07-26 | 2012-08-08 | 日本電気株式会社 | PON system, station side apparatus, and redundancy method used therefor |
JP4807200B2 (en) * | 2006-09-13 | 2011-11-02 | Kddi株式会社 | Optical termination system, optical termination unit and switching method |
US20080195881A1 (en) * | 2007-02-12 | 2008-08-14 | Bernard Marc R | Method and apparatus for providing on-demand backup power for an optical network terminal |
WO2008126162A1 (en) * | 2007-03-06 | 2008-10-23 | Mitsubishi Electric Corporation | Optical communication network system, parent station optical communication device, optical communication method, and communication program |
JP4274265B2 (en) * | 2007-04-27 | 2009-06-03 | 沖電気工業株式会社 | Optical access system and ranging method thereof |
US8817817B2 (en) * | 2007-06-06 | 2014-08-26 | Hewlett-Packard Development Company, L.P. | Method for ethernet power savings on link aggregated groups |
CN101499898A (en) * | 2008-02-03 | 2009-08-05 | 中兴通讯股份有限公司 | Method and apparatus for cipher key interaction |
US8000602B2 (en) * | 2008-04-17 | 2011-08-16 | Pmc-Sierra Israel Ltd. | Methods and devices for reducing power consumption in a passive optical network while maintaining service continuity |
CN101707505B (en) * | 2008-08-13 | 2013-08-28 | 华为技术有限公司 | Method and device for time synchronization in passive optical network and passive optical network |
US8325924B2 (en) * | 2009-02-19 | 2012-12-04 | Microsoft Corporation | Managing group keys |
US8850197B2 (en) * | 2009-07-31 | 2014-09-30 | Futurewei Technologies, Inc. | Optical network terminal management control interface-based passive optical network security enhancement |
-
2010
- 2010-01-31 US US12/697,253 patent/US8422887B2/en active Active
-
2011
- 2011-01-26 JP JP2011013690A patent/JP5570026B2/en active Active
- 2011-01-31 CN CN201110032961.7A patent/CN102256188B/en active Active
-
2013
- 2013-03-12 US US13/794,834 patent/US20130188946A1/en not_active Abandoned
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6975586B1 (en) * | 1999-03-30 | 2005-12-13 | Nec Corporation | Protection switching method and apparatus for passive optical network system |
US20020071148A1 (en) * | 2000-12-12 | 2002-06-13 | Purse Christopher M. | Optical communications network and nodes for forming such a network |
US20040208510A1 (en) * | 2002-03-18 | 2004-10-21 | Katsuichi Ohara | Optical transmission equipment |
US7398018B2 (en) * | 2002-03-18 | 2008-07-08 | Fujitsu Limited | Optical transmission equipment |
US20070058973A1 (en) * | 2005-07-29 | 2007-03-15 | Keiji Tanaka | Optical termination system |
US20090067835A1 (en) * | 2007-09-10 | 2009-03-12 | Charles Chen | Method and apparatus for protection switching in passive optical network |
US20100098407A1 (en) * | 2008-10-21 | 2010-04-22 | Teknovus, Inc. | Method and system for protection switching in ethernet passive optical networks |
US20100166419A1 (en) * | 2008-12-30 | 2010-07-01 | Broadlight, Ltd. | Techniques for protecting passive optical networks |
US20110211837A1 (en) * | 2010-02-26 | 2011-09-01 | Hitachi, Ltd. | Method of Recovery From Sleep State of an Onu in a Pon System Capable of Power Saving |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140093232A1 (en) * | 2012-10-01 | 2014-04-03 | David Hood | Method for protection of multi-wavelength passive optical network |
US8953936B2 (en) * | 2012-10-01 | 2015-02-10 | Telefonaktiebolaget L M Ericsson (Publ) | Method for protection of multi-wavelength passive optical network |
WO2015041651A1 (en) * | 2013-09-19 | 2015-03-26 | Hewlett-Packard Development Company, L.P. | Software defined optical network |
US9900673B2 (en) | 2013-09-19 | 2018-02-20 | Hewlett Packard Enterprise Development Lp | Software defined optical network |
Also Published As
Publication number | Publication date |
---|---|
CN102256188A (en) | 2011-11-23 |
JP2011166758A (en) | 2011-08-25 |
CN102256188B (en) | 2014-03-05 |
US8422887B2 (en) | 2013-04-16 |
JP5570026B2 (en) | 2014-08-13 |
US20110188849A1 (en) | 2011-08-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8422887B2 (en) | System for redundancy in Ethernet passive optical networks (EPONs) | |
TWI430594B (en) | Method and system for protection switching in ethernet passive optical networks | |
US8724453B2 (en) | Method and apparatus for implementing trunk optical fiber protection in ethernet passive optical network (EPON) | |
US9179204B2 (en) | Optical network system | |
US8774621B2 (en) | Communication line switching method, communication apparatus, station-side communication apparatus, communication system, and control unit | |
US10031880B2 (en) | Network device and information transmission method | |
AU2014385716B2 (en) | Link switching method, device, and system | |
EP2966798B1 (en) | Active and standby changeover method, apparatus, device, and system for network device | |
JP4913865B2 (en) | Optical communication network system, master station optical communication device, optical communication method and communication program | |
US9667341B2 (en) | Apparatus and method for protection switching in optical transport network (OTN) suporting ODUflex(GFP) | |
US10009137B2 (en) | Optical communication system, station-side device, subscriber device, and optical communication method | |
KR101357958B1 (en) | System and method for switching pon duplicate path having dual port | |
JP2011135142A (en) | Transmission control system, subscriber-side transmitter, and station-side transmitter | |
CN106982397B (en) | Substation communication network based on return detection and time synchronization safety method thereof | |
JP2007124422A (en) | Optical ring network apparatus | |
JP4692236B2 (en) | OLT switching method, optical termination system, and OLT unit | |
JP5932627B2 (en) | PON system, controller, OSU and ONU | |
WO2012136089A1 (en) | Loop detection method and system for passive optical network | |
JP6236488B2 (en) | Station side optical line termination device, redundant device switching method, and redundant device switching program | |
JP6287404B2 (en) | Station side equipment | |
JP5941024B2 (en) | Optical transmission system, subscriber termination device, and optical signal transmission device | |
JP2014135679A (en) | Terminal-side communication device, station-side device and communication failure restoration method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |
|
AS | Assignment |
Owner name: MICROSEMI ISRAEL STORAGE SOLUTIONS LTD., ISRAEL Free format text: CHANGE OF NAME;ASSIGNOR:PMC-SIERRA ISRAEL LTD.;REEL/FRAME:043379/0152 Effective date: 20160717 |